e-Technology and Manufacturing Enterprise Competitiveness , Volume 17, Issue 6 9781846630811, 9781846630804

Citation preview

jmtm cover (i).qxd

21/07/2006

13:08

Page 1

ISSN 1741-038X

Volume 17 Number 6 2006

Journal of

Manufacturing Technology Management E-technology and manufacturing enterprise competitiveness Guest Editors: Dr Panayiotis H. Ketikidis, Dr S.C. Lenny Koh and Professor Angappa Gunasekaran

www.emeraldinsight.com

Journal of Manufacturing Technology Management

ISSN 1741-038X Volume 17 Number 6 2006

E-technology and manufacturing enterprise competitiveness Guest Editors Dr Panayiotis H. Ketikidis, Dr S.C. Lenny Koh and Professor Angappa Gunasekaran

Access this journal online _________________________

679

Editorial review board ____________________________

680

Guest editorial ___________________________________________

681

Operational intelligence discovery and knowledge-mapping approach in a supply network with uncertainty S.C.L. Koh and K.H. Tan________________________________________

687

Risk management in supply chain: a real option approach Federica Cucchiella and Massimo Gastaldi __________________________

700

Managing stock-outs effectively with order fulfilment systems Richard Pibernik _______________________________________________

721

Optimizing supply chain management using fuzzy approach N. Gunasekaran, S. Rathesh, S. Arunachalam and S.C.L. Koh __________

Access this journal electronically The current and past volumes of this journal are available at:

www.emeraldinsight.com/1741-038X.htm You can also search more than 100 additional Emerald journals in Emerald Fulltext (www.emeraldinsight.com/ft) and Emerald Management Xtra (www.emeraldinsight.com/emx) See page following contents for full details of what your access includes.

737

CONTENTS

CONTENTS continued

A knowledge-based service automation system for service logistics C.F. Cheung, Y.L. Chan, S.K. Kwok, W.B. Lee and W.M. Wang ________

750

The applicability of a multi-attribute classification framework in the healthcare industry Konstantinos Danas, Abdul Roudsari and Panayiotis H. Ketikidis _______

772

Systems and application development for portable maintenance aid (PMA) – a performance perspective Tim S. Leung, Ka Wing Lee and Walter W.C. Chung _________________

786

E-business capabilities model: validation and comparison between adopter and non-adopter of e-business companies in UK Khalid Hafeez, Kay Hooi Keoy and Robert Hanneman ________________

806

An analysis of the relationship between TQM implementation and organizational performance: evidence from Turkish SMEs Mehmet Demirbag, Ekrem Tatoglu, Mehmet Tekinkus and Selim Zaim________________________________________________

829

Transforming the supply chain Evangelia D. Fassoula___________________________________________

848

www.emeraldinsight.com/jmtm.htm As a subscriber to this journal, you can benefit from instant, electronic access to this title via Emerald Fulltext or Emerald Management Xtra. Your access includes a variety of features that increase the value of your journal subscription.

Additional complimentary services available

How to access this journal electronically

E-mail alert services These services allow you to be kept up to date with the latest additions to the journal via e-mail, as soon as new material enters the database. Further information about the services available can be found at www.emeraldinsight.com/alerts

To benefit from electronic access to this journal, please contact [email protected] A set of login details will then be provided to you. Should you wish to access via IP, please provide these details in your e-mail. Once registration is completed, your institution will have instant access to all articles through the journal’s Table of Contents page at www.emeraldinsight.com/1741-038X.htm More information about the journal is also available at www.emeraldinsight.com/ jmtm.htm Our liberal institution-wide licence allows everyone within your institution to access your journal electronically, making your subscription more cost-effective. Our web site has been designed to provide you with a comprehensive, simple system that needs only minimum administration. Access is available via IP authentication or username and password.

Your access includes a variety of features that add to the functionality and value of your journal subscription:

Connections An online meeting place for the research community where researchers present their own work and interests and seek other researchers for future projects. Register yourself or search our database of researchers at www.emeraldinsight.com/ connections Emerald online training services You can also access this journal online. Visit www.emeraldinsight.com/training and take an Emerald online tour to help you get the most from your subscription.

Key features of Emerald electronic journals Automatic permission to make up to 25 copies of individual articles This facility can be used for training purposes, course notes, seminars etc. This only applies to articles of which Emerald owns copyright. For further details visit www.emeraldinsight.com/ copyright Online publishing and archiving As well as current volumes of the journal, you can also gain access to past volumes on the internet via Emerald Fulltext or Emerald Management Xtra. You can browse or search these databases for relevant articles. Key readings This feature provides abstracts of related articles chosen by the journal editor, selected to provide readers with current awareness of interesting articles from other publications in the field. Non-article content Material in our journals such as product information, industry trends, company news, conferences, etc. is available online and can be accessed by users. Reference linking Direct links from the journal article references to abstracts of the most influential articles cited. Where possible, this link is to the full text of the article. E-mail an article Allows users to e-mail links to relevant and interesting articles to another computer for later use, reference or printing purposes. Structured abstracts Emerald structured abstracts provide consistent, clear and informative summaries of the content of the articles, allowing faster evaluation of papers.

Choice of access Electronic access to this journal is available via a number of channels. Our web site www.emeraldinsight.com is the recommended means of electronic access, as it provides fully searchable and value added access to the complete content of the journal. However, you can also access and search the article content of this journal through the following journal delivery services: EBSCOHost Electronic Journals Service ejournals.ebsco.com Informatics J-Gate www.j-gate.informindia.co.in Ingenta www.ingenta.com Minerva Electronic Online Services www.minerva.at OCLC FirstSearch www.oclc.org/firstsearch SilverLinker www.ovid.com SwetsWise www.swetswise.com

Emerald Customer Support For customer support and technical help contact: E-mail [email protected] Web www.emeraldinsight.com/customercharter Tel +44 (0) 1274 785278 Fax +44 (0) 1274 785204

JMTM 17,6

680

EDITORIAL REVIEW BOARD

Harry Boer University of Aalborg, Denmark

Douglas K. Macbeth University of Glasgow, UK

Nourredine Boubekri Northern Illinois University, USA

Bart MacCarthy Business School, University of Nottingham, Nottingham, UK Marly Monteiro de Carvalho Universidade de Sao Paulo, Brazil Les Mitchell University of Hertfordshire, UK Shunji Mohri University of Hokkaido, Japan Andy Neely Cranfield University, UK

Mike Byrne University of Nottingham, UK Felix T.S. Chan The University of Hong Kong, Hong Kong Ian Gibson National University of Singapore, Singapore A. Gunasekaran University of Massachusetts, USA Abdel-Aziz Hegazy Helwan University, Egypt Bob Hollier Manchester Business School, UK Hiroshi Katayama Waseda University, Japan Tarek Khalil University of Miami, USA Ashok Kochhar University of Aston, UK Siau Ching Lenny Koh University of Sheffield, UK Doug Love University of Aston, UK

Journal of Manufacturing Technology Management Vol. 17 No. 6, 2006 p. 680 # Emerald Group Publishing Limited 1741-038X

Kul Pawar University of Nottingham, UK Roy Snaddon University of Witwatersrand, South Africa Amrik Sohal Monash University, Australia Mile Terziovski The University of Melbourne, Australia Wang Xing Ming Renmin University, China Peter Wright Celestica, UK

Guest editorial

Guest editorial

A special issue on e-technology and manufacturing enterprise competitiveness 681 About the Guest Editors Dr Panayiotis Ketikidis is the Vice Principal of CITY Liberal Studies – an affiliated institution of the University of Sheffield and is responsible for the Computer Science department and the Centre for Information Services. His main role within the South East European Research Centre (SEERC) is the overall management of the centre and the development and implementation of SEERC future research agenda. Also to provide strategic leadership in the planning of SEERC research projects and to assist in strengthening existing partnerships and establishing new ones in the South East European (SEE) region as well as with government, non-governmental organisations and with industry and the private sector. Dr Ketikidis’ early work was with leading research institutes in the USA. He is also the President for the International Society of Logistics – Chapter Thessaloniki (1999-2005). He has also been involved in many European Information Technology Projects funded by the European Commission (1992-current) and he is a member of organising and scientific committees in various national/international conferences/workshops and of the editorial board of the International Journal of Operational Research. His research interests are: ICT policy, logistics information systems, medical informatics and research strategy. Dr S.C. Lenny Koh is the Director of the Logistics and Supply Chain Management Research Group and an Associate Professor/Senior Lecturer in Operations Management at the University of Sheffield Management School, UK. She holds a doctorate in Operations Management and a first-class honours degree in Industrial and Manufacturing Systems Engineering. Her research interests are in the areas of production planning and control (ERP and ERPII), uncertainty management, modern operations management practices, logistics and supply chain management, e-business, e-organisations, knowledge management, sustainable business and eco-logistics. Dr Koh has 166 publications in journal papers, book, edited book, edited proceedings, edited special issues, book chapters, conference papers, technical papers and reports. She is the Editor-in-Chief of the International Journal of Enterprise Network Management and the Associate Editor of the International Journal of Operational Research. She is on the editorial board of several international journals and has guest edited many high profile journals. She organised and chaired international conferences and on the board of scientific/international/programme committee of many international conferences. She has received grants and awards from several national and international funding bodies, and has been a consultant to SMEs and large enterprises. Dr Angappa Gunasekaran is the Director of Business Innovation Research Center and a Professor of Operations Management in the Department of Management at the Charlton College of Business, University of Massachusetts (Dartmouth, USA). He is teaching undergraduate and graduate courses in operations management and management science. Dr Gunasekaran has over 175 articles published in 40 different peer-reviewed journals. He has presented about 50 papers and published about 50 articles in conferences and given a number of invited more talks in about 20 countries. Dr Gunasekaran is on the Editorial Board of over 20 journals. He has organized several international workshops and conferences in the emerging areas of operations management and information systems. Dr Gunasekaran has edited a couple of books. He is the Editor of Benchmarking: An International Journal. He has edited special issues for a number of highly reputed journals. Dr Gunasekaran is currently interested in researching benchmarking, management information systems, e-commerce (B2B), information technology/systems evaluation, performance measures and metrics in new economy, technology management, logistics, and supply chain management. He actively serves on several university committees.

Journal of Manufacturing Technology Management Vol. 17 No. 6, 2006 pp. 681-686 q Emerald Group Publishing Limited 1741-038X

JMTM 17,6

682

This special issue of Journal of Manufacturing Technology Management (JMTM) on “e-Technology and Manufacturing Enterprise Competitiveness” mainly comprises of selected papers presented at the 3rd International Workshop on Supply Chain Management and Information Systems (SCMIS2005), hosted by the SEERC – a collaborative venture of the University of Sheffield (UK) and Thessaloniki’s CITY Liberal Studies (Greece) – affiliated institution of the University of Sheffield, on 6-8 July 2005, in Thessaloniki, Greece. This workshop was jointly organised by the University of Sheffield (UK), CITY Liberal Studies – affiliated institution of the University of Sheffield (Greece), SEERC (Greece), University of Nottingham (UK), The Hong Kong Polytechnic University (China), University of Massachusetts (USA), Aristotle University of Thessaloniki (Greece), University of Macedonia (FYROM), Athens University of Economics and Business (Greece), and National Chung Hsing University (Taiwan). Globalisation, modernisation and streamlining paradigms have driven many enterprises to use various e-technologies in order to improve the performance of existing operations, and compete globally and strategically to enhance manufacturing enterprise competitiveness, which in today’s digital economy, is often networked and interconnected via the internet, intranet, and extranet. Examples of the e-technology include e-commerce, e-business, e-procurement and e-logistics. These technologies are in place to support the notion of establishing a value-added e-supply chain and e-demand chain. The support of back-office systems, e.g. supply chain management, manufacturing resource planning, enterprise resource planning are crucial to enable seamless information flow in the supply chain, whilst support from front-office systems, e.g. customer relationships management, is important to coordinate the demand chain. Appropriate alignment of the e-technology with the systems is expected to create further competitive advantages. Hence, e-technology is a core competence in contributing to competitiveness in the digital economy. It is not merely a facilitating enabler, but a critical enabler towards globalisation. Past research on aspects of e-technology tends to be fragmented. In today’s digital economy, practitioners of production and operations need to address their evolving business landscape in formulating, designing and implementing competitive solutions in order to add value to their business. This special issue aims to investigate the performances, successes, failures, critical issues and roles of e-technology in contributing to manufacturing enterprise competitiveness. The special issue consists of theory building and empirical study papers that have strong relevance to the practical world in manufacturing sectors, as well as case studies and papers with best practices experiences. This special issue contains ten papers contributed by researchers and practitioners from Hong Kong (China), India, Italy, Greece, Spain, Turkey, UK and USA. The papers cover a diverse range of contributions including research on operational intelligence discovery, risk and uncertainty in supply chains, total quality management (TQM) related performance, supply chain transformation; and application of order fulfilment systems, knowledge mapping and knowledge-based approach, real option approach, fuzzy approach, classification framework, portable maintenance system and e-business capabilities model. The normal JMTM review guidelines were followed.

Koh and Tan in their paper, “Operational intelligence discovery and knowledge mapping approach in a supply network with uncertainty” propose an approach for discovering operational intelligence and enabling knowledge mapping in a supply network with uncertainty. Knowledge mapping and handbook techniques were used and operationalised with a software tool named TAPS. TAPS was used to model a supply network with uncertainty and to discover operational intelligence in a supply network. This research found that knowledge management itself is inadequate for managing a supply network with uncertainty. It should be supplemented by knowledge mapping with capability of identifying the operational intelligence within the supply network. To this end, iTAPS was developed in order to provide managers with an ability to visualise the operational intelligence for a given objective, and to identify the likely effects on implementing a particular tool or technique in a supply network. The main contribution from this research was the proposed new approach called the “intelligence handbook” used to discover operational intelligence in order to map knowledge in a supply network with uncertainty. The paper, “Risk management in supply chain: a real option approach” by Cucchiella and Gastaldi, aims to individualise a framework for the management of uncertainty in supply chain aiming to reduce firm risks. They argued that a way for reducing the damages derived from uncertainty sources is through increasing the level of flexibility within the supply chain. The real option theory would enable the addition of flexibility in the supply chain. They applied the real options theory to study risks within the supply chain. A theoretical framework has been individualised, enabling the selection of possible options to protect the firm against the risk originating from sources of uncertainty. Using Matlab for testing, it was found that the ability of outsource option can minimise firm risks. Pibernik in his paper, “Managing stock-outs effectively with order fulfilment systems” provides a consistent formal approach to modelling order promising mechanisms, introduces new and innovative order promising mechanisms and gives valuable insight into their performance through numerical analysis. In this research, a different order promising mechanism was suggested, and analysed to examine how well this mechanism could contribute to the effective management of stock-out situations. A formal description and analysis of alternative order promising mechanism applicable in make-to-stock systems was provided. This empirical research was conduced based on the data of a pharmaceutical company. The potential of alternative order promising mechanisms to alleviate the negative consequences associated with a temporary stock-out situation were noted. The results obtained from this analysis provide guidelines for manufacturers, retailers, and vendors of supply chain software on how to design and utilise order promising systems. The paper, “Optimising supply chain management using fuzzy approach” by Gunasekaran, Rathesh, Arunachalam and Koh, put forward a fuzzy multi-criteria decision-making procedure and it was applied to find a set of optimal solution with respect to the performance of each supplier in a supply chain. Responses obtained from customers were simulated using a triangular fuzzy quality function deployment (QFD) algorithm, Monte Carlo simulation and a multi-objective model to optimise the total user preferences. It was identified that this method with the use of Monte Carlo simulation produces overall desirability level less imprecise and more realistic than those of the conventional QFD methods for engineering design evaluation.

Guest editorial

683

JMTM 17,6

684

Cheung, Chan, Kwok, Lee and Wang in their paper, “A knowledge-based service automation system for service logistics” highlight that effective service logistics can lower the cost and increase service value by improving customer satisfaction and loyalty. It was purported that although the concept of service logistics has been used in different sectors of the industries, the advantages of service logistics have been overlooked by high-tech and high value manufacturing industries. This paper presents a knowledge-based service automation system (KBSAS) to enhance the competitiveness for manufacturing enterprises in service logistics. The KBSAS incorporates various artificial intelligence technologies such as case-based reasoning which is used for achieving four perspectives of knowledge acquisition, service logistics, service automation and performance measurement, respectively. A prototype customer service portal has been built based on the KBSAS and implemented successfully in a semiconductor equipment manufacturing company. The paper, “The applicability of a multi-attribute classification framework in the healthcare industry” by Danas, Roudsari and Ketikidis, introduces a new approach to address the inefficiencies of hospital pharmacy management. The study examines the applicability of the Ned-MASTA classification method for managing medicines inventory within the hospital pharmacy environment. A virtual pharmacy inventory system was suggested that forms a virtual pharmacy inventory of hospitals within the same geographical region. These facilities were envisaged to provide the infrastructure for the cooperation of hospital pharmacies in order to improve the efficiency of their operations. A survey was conducted among Greek hospitals to identify the inefficiencies of their logistics systems. They argued that it was considered vital and necessary to investigate the solutions adopted by other industries that are facing similar problems. Hence, the case of spare parts inventory for production machines was used to present similarities with the management of medicine stock within the hospital pharmacy. The approach that was adopted by the spare parts case was adapted and included in the system that forms a virtual hospital pharmacy inventory. Then, the MASTA classification approach was modified to fit in the operation of the hospital pharmacy and a system was constructed to form the virtual pharmacy inventory. The applicability of the system was demonstrated through an application scenario using a prototype system. Leung, Lee and Chung in their paper, “Systems and application development for portable maintenance aid (PMA) – a performance perspective” explores the theory of technology adoption using a system-application approach to facilitate PMA innovation span across multiple organisations in order to enhance overall production network effectiveness. The survey on available technology and forecast future trends led to the development of a workable prototype for testing in real-world airline operation under a three-layer analysis model. Industrial data was used to validate the concept and correlate performance factors in PMA adoption. This research revealed that greater operational efficiency and higher process value maintenance can be achieved through communication paths. The value process-to-shop concept was validated, next to the value network. This research provided maintenance business, especially with high contents of information and knowledge driven tasks for outsourcing service providers, with the concept of the adoption rules for wireless system for assets and maintenance performance management.

The paper, “e-Business capabilities model: validation and comparison between adopter and non-adopter of e-business companies in UK” by Hafeez, Keoy and Hanneman, presents a conceptual framework to evaluate e-business strategic capabilities using structural equation modelling approach. Three e-business capabilities namely business strategy, supply chain strategy and e-business readiness were identified. These capabilities were further categorised under technology, organisation and people tracks/paths/branches to assess their contribution in business effectiveness. Survey data from 143 firms from the UK was collected to test the theoretical model. They found a positive, mediating/reciprocal relationships among multidimensional measures of business strategy, supply chain strategy and e-business adoption. Their empirical analyses demonstrated several key findings: . Success of e-business in UK firms is attributed to the strong positive correlation of supply chain strategy to business strategy and to e-business adoption. . Within the technology-organisation-people dimensions, e-business adoption and business strategy emerge as the strongest factors for the company’s performance for the adopter of e-business group, whereas supply chain capabilities and business strategies are relatively stronger contributory factors towards business success for non-adopter of e-business. It was envisaged that these findings would provide useful guidelines for companies to assess their strengths and weaknesses towards adopting an effective e-business implementation strategy. Demirbag, Tatoglu, Tekinkus and Zaim in their paper, “An analysis of the relationship between TQM implementation and organisational performance: evidence from Turkish SMEs” determines the critical factors of TQM and measures their effect on organisational performance drawing on a sample of 141 SMEs operating in the Turkish textile industry. Using exploratory and confirmatory factor analyses, seven empirically validated dimensions of TQM were identified. Organisational performance was measured using subjective measures relying on executive’s perception of how the firm performed relative to both financial and non-financial performance criteria. The structural equation modelling technique was employed to investigate the relationship between the implementation of TQM practices and organisational performance. Data analysis revealed that there is a strong positive relationship between TQM practices and non-financial performance of SMEs, while there is only weak influence of TQM practices on financial performance of SMEs. With only a mediating effect of non-financial performance, TQM practice has a strong positive impact on financial performance of SMEs. The final paper, “Transforming the supply chain” by Fassoula analyses the transformation process of supply chain and provides a modular structured management tool for planning, implementing and measuring the effectiveness of supply chain transformation process, in relation to overall organisational performance and business strategy. It was purported that overall organisational performance is meant to reflect the satisfaction rate of all interested parties (customers, employees, stakeholders, suppliers and social partners) and the reliability of operations. The critical sensitivity factors of a transformation process formed the conceptual platform for the proposed management tool.

Guest editorial

685

JMTM 17,6

686

We could not have done this by ourselves and we totally appreciate the efforts and support of all who were involved in making this special issue possible, which includes the authors, referees, Editor of JMTM – Professor David Bennett, and the editorial staff of JMTM. The guest editors gratefully acknowledge the assistance provided by the Editor of JMTM and the referees who reviewed the manuscripts for this special issue. Panayiotis H. Ketikidis South East European Research Centre, CITY Liberal Studies – Affiliated Institution of the University of Sheffield, Thessaloniki, Greece S.C. Lenny Koh University of Sheffield, Management School, Sheffield, UK, and Professor Angappa Gunasekaran Department of Management, Charlton College of Business, University of Massachusetts, North Dartmouth, Massachusetts, USA

The current issue and full text archive of this journal is available at www.emeraldinsight.com/1741-038X.htm

Operational intelligence discovery and knowledge-mapping approach in a supply network with uncertainty S.C.L. Koh

Operational intelligence

687 Received October 2005 Revised December 2005 Accepted January 2006

University of Sheffield, Management School, Sheffield, UK, and

K.H. Tan Nottingham University Business School, Nottingham, UK Abstract Purpose – The purpose of this research is to propose an approach for discovering operational intelligence and knowledge mapping in a supply network with uncertainty. Design/methodology/approach – Knowledge mapping and handbook techniques are used. TAPS software is used to model a supply network with uncertainty and to discover operational intelligence in a supply network. Findings – Knowledge management is inadequate for managing a supply network with uncertainty. Knowledge mapping is proposed, but it needs to be assisted by operational intelligence. Practical implications – iTAPS provides managers with an ability to visualise the operational intelligence for a given objective, and to identify the likely effects on implementing a particular tool or technique in a supply network. Originality/value – A new approach – called the “intelligence handbook” is proposed to discover operational intelligence in order to map knowledge in a supply network with uncertainty. Keywords Knowledge management, Knowledge mapping, Uncertainty management, Supply chain management, Decision making Paper type Research paper

1. Background The performance of today’s global supply network is affected by an increased number of uncertainties. Koh and Saad (2004) defined uncertainty as any unpredictable event that affects the performance of an enterprise. In an economy where the only certainty is uncertainty, the one sure source of lasting competitive advantage is knowledge (Nonaka, 1998). By taking a “zoom-in view” of a simple example: a supply network that consists of a manufacturer from China and a distributor in the UK, it can be noted that various types of uncertainty may affect the performance of such a supply network. For instance, the logistics issues on ability to deliver on-time, the expected quality in the UK, and others. It is clear that the setting up of a supply network process is connected to the presence of operational interdependencies between the units of the supply chain, the stability and effectiveness of a supply network is closely bound to the ability of the core firm to plan

Journal of Manufacturing Technology Management Vol. 17 No. 6, 2006 pp. 687-699 q Emerald Group Publishing Limited 1741-038X DOI 10.1108/17410380610678747

JMTM 17,6

688

the governance structure of the supply relationship, and the product structure and the nature of the process influence the networking process (De Toni and Nassimbeni, 1995). This phenomenon calls for a better understanding of the operations and characteristics of a supply network with uncertainty, so that a more accurate knowledge can be mapped and a better decision can be made. Following with this notion, much past research has been carried out proposing various methods of knowledge management in a supply chain. Thannhuber et al. (2001) developed an autopoietic approach for building knowledge management systems in manufacturing enterprises. It was suggested that through understanding of the knowledge at the enterprise level, one could cope with dynamic changes of market demand, process design, capabilities in diverse locations, and associated fluctuations in internal process adaptations. Their work has also indicated that instead of focusing on individual human knowledge, the ability of an enterprise to dynamically derive processes to meet the external needs and internal stability is identified as the organisational knowledge. It was proposed that this autopoietic approach could be used in managing a supply network. Towill (1996) suggested that an industrial dynamics modelling of real-life supply chains is a powerful methodology for predicting and prioritising methods of reengineering the chain to achieve enhanced performance when viewed from the perspective of all players in the chain. However, building an adequate model of an existing or proposed real-life supply chain requires the use of people-based resources, observation-based resources and systems knowledge-based resources. Therefore, it is important to be able to map the knowledge of a supply chain particularly in a global supply network with uncertainty. Automotive multinational firms increasingly engage in transferring their supply management practices across countries within their global network of operational units. Pagano (2003) provided an in-depth qualitative analysis on such transfer outcomes and on relevant factors shaping such processes. The case of a component automotive manufacturer in China was investigated. The results showed that the transfer process has concerned mainly the supply chain decision-making procedures and those areas linked to the production phase including delivery/logistics and quality control. Using the concept of knowledge transfer, Graziadio and Zilbovicius (2003) compared the supply system in two plants of a single car assembler: one is modular and the other is conventional, both producing sub-compact cars. The modular plant has the highest level of outsourcing. Modular supply demands more interaction between assembler and systemist (also called modular supplier) whenever the logic of outsourcing is present. As long as the assembler transfers more responsibilities for design, purchasing and production to the systemists, the flow of information has to be very efficient. The volume and intensity of knowledge transferred from assembler to systemist also depends on the systemist’s role. Using the concept of knowledge sharing, it was argued that supplier performance will benefit most where time-bound relational assets have developed between a buyer and supplier and the firms exploit the resulting communication, efficiency by transferring productive knowledge (Kotabe et al., 2003). In that study, the effects of two forms of knowledge exchange were also examined, together with the prior duration of the buyer-supplier relationship. Similar interaction patterns were found in two survey samples of Japanese and the US automotive suppliers. Nath et al. (2005) studied the supply chain management of two automobile manufacturers in India and traced the relationship between supply chain and

knowledge chain management. The efforts of building up the supply chain were examined under three broad headings namely, knowledge hierarchy, geographical proximity and cultural reorientation of the supply chain. An enterprise’s knowledge domain, defined as critical, sub-critical or common market knowledge, was related to various organisational forms. As the North American automotive manufacturers transfer more responsibility to major suppliers while at the same time experiencing significant engineering personnel reductions due to layoffs and retirements, the industry landscape begins to change (Belzowski et al., 2003). It was recommended that companies must understand thoroughly the value of knowledge within the organisation; acknowledge likely gaps between the perceived benefits and reported knowledge activity levels; resolve discontinuities in knowledge sharing activities between the company, its customers and suppliers; take into account differing emphases by the company, its customers and suppliers on people, technology, process and culture as facilitators of knowledge activities; and measure knowledge activities in order to manage them effectively. It must be noted that increasing globalisation of aerospace supply network could also be identified. This is particularly true in the outsourcing of engineering and manufacturing operations. In this type of outsourcing project, the effective use of knowledge across the supply chain is crucial. Fan et al. (2000) studied the different types of supplier knowledge in the five key stages of the aerospace product development process, and used an existing joint supplier improvement information system to illustrate best practice in supply base management. Outsourcing has not only dominated the manufacturing industry, but has equally dominated the service industry. For examples, outsourced of call centres and software development to India, are amongst the most practised operations. Koh et al. (2004) developed a knowledge management model for effective service performance for a call centre. Such operation also requires an effective knowledge management procedure in order to protect innovations capability and service performance. 2. Research gaps, aims and objectives It can be synthesised from the above literature review that little research has been carried out on knowing mapping. Simply identifying and managing the relevant knowledge in an enterprise or supply network will not be suffice because of the inter-relations of business functions and operations in a supply network, particularly one with uncertainty and with many tiers. Without proper mapping of knowledge, the causes and effects of uncertainty in a supply network may not be clearly understood, and consequently this may lead to inaccurate decision-making. Knowledge mapping is difficult when there is uncertainty in a supply network. Owing to its unpredictable nature, which complicates the process of knowledge mapping, it is necessary to extract the operational intelligence within a supply network so that the operational intelligence can be used as “clues” for predication on probable uncertainty. The review above shows little work on analysing “knowledge” with “intelligence”. In this research, operational intelligence can be defined as a kind of integrated knowledge at a level higher than tacit knowledge on operations within a supply network. For example, the time for repairing a machine, which is not recorded and is a form of experience of the operator, can be deemed a type of tacit knowledge. Integrating this tacit knowledge to other related operations in a supply network, it may

Operational intelligence

689

JMTM 17,6

690

inform the decision maker that a probable delay due to machine breakdown can be minimised with the use of a different spare part. Thus, this operational intelligence activates an informed decision-making, which attempts to deal with uncertainty by integrating knowledge from other sources in a supply network. To enable knowledge mapping, discovery of operational intelligence is necessary. Brooks (1991) noted that an intelligent system is decomposed into independent and parallel activity producers which all interface directly to the world through perception and action, rather than interface to each other particularly much. Business intelligence is also an important source of operational intelligence. Anon (2000) suggested that evolving applications will support a closed loop decision-making process whereby the output of business intelligence applications is routed to operational system users in the form of suggested actions which could be taken to remedy specific business problems. Distributed intelligence has been proposed by Lu¨der and Peschke (2004) as a flexible approach for controlling plant automation. Human intelligence has been suggested by Zhou and Chuah (2000) as the key factor for successful intelligent manufacturing. These are possible sources of operational intelligence to assist knowledge mapping and uncertainty management. Garavelli and Gorgoglione (2003) proposed a framework to analyse manufacturing problems according to a knowledge management perspective. The analysis is based on the main knowledge processes involved in problem solving, namely knowledge generation, memory, transfer and codification. Each can be assessed by variables such as uncertainty, space-dependence, time-dependence, and codification level, describing manufacturing problems in terms of, for instance, degree of coordination and integration, repetitiveness or uniqueness of the solution. Owing to multi-tiers of suppliers in a supply network, codification and knowledge sharing may not be straight forward. This research proposes an approach for discovering operational intelligence and knowledge mapping in a supply network with uncertainty. The approach is developed by referring to the concept of new value stream or supply chain mapping (Hines and Rich, 1997) and by adapting the notion of using operational intelligence as “clues” for uncertainty management. It is envisaged that this approach will assist better decision-making. 3. Knowledge-mapping techniques The knowledge embodied in managers’ experience is tacit. Ideally, managers need a knowledge-mapping tool that gives them a way of eliciting and capturing this knowledge, a mechanism for retaining it and, if possible, a way of providing a comprehensive check. A typical output could be a formal explicit model of variable linkages, in contrast to their previous tacit mental models. Once the variable linkages are understood, analysis is required to determine the potential impact of changes being considered. A knowledge-mapping tool that automates this analysis would allow managers to consider a wide range of options in a short time. The final stage, an evaluation of the options, requires the consideration of many factors, so some form of multi-attribute decision-making is required. A knowledge-mapping tool should provide the appropriate level of functionality and detail, yet be easy to use. In short, the requirements of a tool for knowledge mapping are (Tan and Platts, 2002): . Sequential decision-making. Supports managers through the entire process from identifying relevant variables to evaluating decisions.

Visualization support. Provides visualization support at each stage of the decision-making process. . Integrated documentation. Captures information on variables and linkages for analysis or comparison. 3.1 Appraisal of existing techniques When setting out on the task of knowledge mapping and operational intelligence gathering, managers have available to them several existing techniques to generate ideas, and structure and analyse problems. Could these techniques sufficiently help managers to identify a range of actions? In order for managers to generate a wide range of actions they need to identify the relevant variables within a problem situation, to develop an understanding of these variables and the linkages among them, to analyse these linkages and, hence, identify actions, tools and techniques that they can use. Finally, the alternatives need to be evaluated and an action plan to be compiled for further implementation. Tan and Platts (2003) made a comparison of existing causal mapping techniques for knowledge mapping (Figure 1). They pointed out that the general purpose mapping tools described above could be used for the first part of this task. They provide a way of scoping a problem and identifying relevant variables. However, because they are general purpose, they are not necessarily optimized for the knowledge-mapping task. For example, cognitive mapping might result in overly complex models since it allows the development of multiple foci, whereas fishbone diagrams, created for specific problems with clear boundaries, might be too simplistic. A range of commercially available software packages has been built around the techniques discussed above. These software tools (Figure 1) automate the application of the techniques and enhance information visualization. However, they generally address a specific stage of decision-making process and do not provide comprehensive support through all the stages. This work aims to provide managers with a software tool that supports their analysis at every stage of the decision-making process. .

4. An “intelligence handbook” approach Tan and Platts (2003) developed a software tool called TAPS. The tool is implemented under the Microsoft Windows operating system using Microsoft’s Visual Basic 6.0 programming language. TAPS has four main modules: (1) database; (2) graphic user interface; (3) algorithm; and (4) evaluation. They use a network diagram to represent the inter-relationship between a variable and its connected variables. In the network diagram the variable is displayed as a node with edges (lines) linking it to other nodes with which it has a connectance. Arrows connect variables to indicate the existence and direction of a connectance. Thus, a variable’s connectance network is made up of nodes and relations. Having studied and investigated the TAPS approach in details, we believe that it could be used as a tool to address the issues of knowledge-mapping and intelligence gathering in a supply network.

Operational intelligence

691

JMTM 17,6

692

Figure 1. Comparison of causal mapping techniques

A TAPS network diagram has five basic levels. The bottom level displays the objective or the variable on which analysis is to be carried out. At level two, the objective is broken down into its different dimensions. For example, “Increase flexibility” can be broken into four resource dimensions (Slack et al., 1995): System flexibility, labour flexibility, process flexibility, and control flexibility. At level three, the relevant cause-effect variables for each dimension are displayed. The fourth level displays the actions that could be taken to address the variables. For example, the variables affecting labour flexibility could be training and working hours. One of the actions that could be taken to address “working hours” is overtime (Figure 2). At level five are tools that could be used to address particular actions or variables. To enable a user to identify actions related to specific variables, TAPS has built-in functions called “trace-down” and “trace-up”. Trace-down is an analysis to determine the effect of a given direction of change in one variable on other variables in the network. Trace-up analysis, starts with a desired direction of change in a variable and works back to find which other variables have to be changed in value and in which direction. These functions allow a user to perform an analysis on any variable in the network diagram. Please see Tan and Platts (2003) for further descriptions of TAPS functions.

Operational intelligence

693

4.1 Mapping operational intelligence Tan and Platts (2002) argued that there is no single or best way of categorising management tools and techniques and categorisations will depend on the task at hand. For this task the main requirements are: . Comprehensiveness. The handbook should cover all known techniques, tools and variables. . Linkages. The handbook should indicate to managers the inter-relationships among tools, techniques and variables. . Visualisation. The inter-relationship among tools, techniques and variables should be graphically illustrated. Taking each of these in turn. SMED

Reduce Setup Time

Overtime

Training

System

Zero Defect

Workhour

New Machine

Capacity

Process

Labor

Tools

Actions

…

…

Control

Variables

Dimensions of objective

Objective Flexibility

Figure 2. Structure of the network diagram for flexibility

JMTM 17,6

.

.

694

.

Comprehensiveness. Clearly it is unlikely that the handbook will include all known techniques. The intention is to allow managers to map out operational information and knowledge in a supply network. Linkages. The existence of inter-relationship among tools, techniques and variables can be shown graphically by placing the names of the tools, techniques or variables within circular nodes and connecting those two nodes with an arrow. Visualisation. The five basic levels of TAPS network shows the inter-relationship between tools, techniques and variables. By using the database functions, managers could codify the source of information from various partners in a supply network.

Figure 3 shows a structure of a handbook. Basically, for a given objective, the handbook could perform three main functions to: Techniques

Techniques

Tools

Tools

Actions

Actions

Actions

Variables

Variables

Tools

Variables

Objective

Actions

Techniques

Variables

Tools

Information Files

Actions

Techniques

Objective

Figure 3. Structure of the handbook

Variables

Operational Intelligence

Information Documents

. .

.

provide an overall view of relevant variables; enable a more detailed view of the operational intelligence (actions, tools and techniques); and; provide information about each tool and technique[1].

We called the operational intelligence approach as Intelligence TAPS (iTAPS).

Operational intelligence

695

5. Application Let us take a look at some examples to illustrate the application of iTAPS. If a manager wants to improve manufacturing flexibility, a trace-up analysis on the “process” dimension of a “flexibility” variable could be conducted. Figure 4 shows the network diagram for a “process” dimension of flexibility. At the upper level are the action plans (actions, tools and techniques). The network illustrates to managers the indirect impact of an action could have on a network. Managers also could use iTAPS to identify suitable actions for managing the supply chain uncertainty. Its suitability is decided based on satisfactory on-time delivery performance achieved by certain cluster of enterprises. The identified alternative actions are then prioritised using the AHP built-in decision support feature of TAPS (Figure 5). Supply chain uncertainty that affects on-time delivery could be tackled via a number of different actions. For example, the possible actions (alternatives) include total preventive maintenance (TPM), training need analysis (TNA), and overall equipment efficiency (OEE) (Figure 6). Three criteria are used to assess the merits of the identified actions in achieving the objective of on-time delivery.

Figure 4. Trace-Up network diagram for “process” dimension (partial view)

JMTM 17,6

696

Figure 5. A partial TAPS model

These include time effectiveness (time taken to provide results), sustainability (continuing for a long time), and integration (ability to integrate well with existing practices). In the above example, the results of the pair-wise comparison suggested that TNA has a major impact on time effectiveness, sustainability and integration. Therefore, TNA could be taken as an action to achieve on-time delivery. 6. Conclusions and implications In this paper we have demonstrated iTAPS, an operational intelligence approach to managing supply chain uncertainty. In practise, iTAPS application is best done in a group. iTAPS provides clarity by facilitating communication among managers and helps to get everyone’s view out in the open, where assumptions can be recognised and challenged and where the need for more detail information becomes more obvious. We believe iTAPS provides managers with following benefits: . an ability to visualise the operational intelligence for a given objective; and . an ability to identify the likely effects on implementing a particular tool or technique in a supply network. In the longer view, iTAPS should facilitate the accumulation and integration of a body of production knowledge, so that what is learned through coping with one set of problems can be brought to bear on others.

Operational intelligence

697

Figure 6. Prioritisation of actions

Note 1. For the intelligence handbook approach, we categorised tools and techniques separately. Tools as ways to carry out a particular task, whereas techniques are programmes or procedures which comprise a set of tools (Tan and Platts, 2002). References Anon, A. (2000), “Business intelligence: getting the most from your ERP systems”, Manufacturing Computer Solutions, Vol. 6 No. 11, pp. 18, 19, 21. Belzowski, B.M., Flynn, M.S., Richardson, B.C., Sims, M.K. and VanAssche, M. (2003), “Harnessing knowledge: the next challenge to inter-firm cooperation in the North American auto industry”, International Journal of Automotive technology and Management, Vol. 3 Nos 1/2, pp. 9-29. Brooks, R.A. (1991), “Intelligence without representation”, Artificial Intelligence, Vol. 47 No. 1, pp. 139-59. De Toni, A. and Nassimbeni, G. (1995), “Supply networks: genesis, stability and logistics implications – a comparative analysis of two districts”, Omega, Vol. 23 No. 4, pp. 403-18. Fan, I.-S., Russell, S. and Lunn, R. (2000), “Supplier knowledge exchange in aerospace product engineering”, Aircraft Engineering and Aerospace Technology, Vol. 72 No. 1, pp. 14-17.

JMTM 17,6

698

Garavelli, A.C. and Gorgoglione, M. (2003), “A knowledge framework for problem solving in the worldwide manufacturing competition”, International Journal of Automotive Technology and Management, Vol. 3 Nos 3/4, pp. 234-48. Graziadio, T. and Zilbovicius, M. (2003), “Knowledge transfer through the supply system: does modularity make it easier?”, International Journal of Automotive Technology and Management, Vol. 1 No. 2, pp. 47-60. Hines, P. and Rich, N. (1997), “The seven value stream mapping tools”, International Journal of Operations & Production Management, Vol. 17 No. 1, pp. 46-64. Koh, S.C.L. and Saad, S.M. (2004), “Modelling uncertainty under a multi-echelon ERP-controlled manufacturing system”, Journal of Manufacturing Technology Management, Vol. 15 No. 3, pp. 239-53. Koh, S.C.L., Gunasekaran, A., Thomas, A. and Arunachalam, S. (2004), “The application of knowledge management in call centres”, Journal of Knowledge Management, Vol. 9 No. 4, pp. 56-69. Kotabe, M., Martin, X. and Domoto, H. (2003), “Gaining from vertical partnerships: knowledge transfer, relationship duration, and supplier performance improvement in the US and Japanese automotive industries”, Strategic Management Journal, Vol. 24 No. 4, pp. 293-316. Lu¨der, A. and Peschke, J. (2004), “Distributed intelligence for plant automation based on multi-agent systems: the PABADIS approach”, Production Planning & Control, Vol. 15 No. 2, pp. 201-12. Nath, P., Sandhya, G.D. and Mrinalini, N. (2005), “Supply chain as knowledge management”, International Journal of Logistics Systems and Management, Vol. 2 No. 3, pp. 267-78. Nonaka, I. (1998), “The knowledge creating company”, Harvard Business Review, special issue on Knowledge Management, July/August, pp. 21-46. Pagano, A. (2003), “The development of global supply management capabilities in the automotive industry: the transfer of supply management practices in the People Republic of China”, International Journal of Automotive Technology and management, Vol. 3 Nos 1/2, pp. 84-100. Slack, N., Chamber, S., Harland, C., Harrison, A. and Johnston, R. (1995), Operations Management, Pitman, London. Tan, K. and Platts, K. (2002), “Managing manufacturing action plans”, International Journal of Innovation Management, Vol. 6 No. 4, pp. 369-86. Tan, K.H. and Platts, K. (2003), “Linking objectives to action plans: a decision support approach based on the connectance concept”, Decision Sciences Journal, Vol. 34 No. 3, pp. 569-93. Thannhuber, M., Tseng, M.M. and Bullinger, H-J. (2001), “An autopoietic approach for building knowledge management systems in manufacturing enterprises”, Annals CIRP, Vol. 50 No. 1, pp. 313-8. Towill, D.R. (1996), “Industrial dynamics modelling of supply chains”, Logistics Information Management, Vol. 9 No. 4, pp. 43-56. Zhou, Y. and Chuah, K.B. (2000), “Human intelligence: the key factor for successful intelligent manufacturing”, Integrated Manufacturing Systems, Vol. 11 No. 1, pp. 30-41. About the authors S.C.L. Koh is the Director of the Logistics and Supply Chain Management Research Group and an Associate Professor/Senior Lecturer in Operations Management at the University of Sheffield Management School UK. She holds a Doctorate in Operations Management and a first-class honours degree in Industrial and Manufacturing Systems Engineering. Her research interests are

in the areas of production planning and control (ERP and ERPII), uncertainty management, modern operations management practices, logistics and supply chain management, e-business, e-organisations, knowledge management, sustainable business and eco-logistics. Dr Koh has 166 publications in journal papers, book, edited book, edited proceedings, edited special issues, book chapters, conference papers, technical papers and reports. She is the Editor in Chief of the International Journal of Enterprise Network Management and the Associate Editor of the International Journal of Operational Research. She is on the editorial board of several international journals and has guest edited many high profile journals. She organised and chaired international conferences and on the board of scientific/international/programme committee of many international conferences. She has received grants and awards from several national and international funding bodies, and has been a consultant to SMEs and large enterprises. S.C.L. Koh can be contacted at: [email protected] K.H. Tan is a lecturer at Nottingham University Business School, Nottingham, UK

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

Operational intelligence

699

The current issue and full text archive of this journal is available at www.emeraldinsight.com/1741-038X.htm

JMTM 17,6

Risk management in supply chain: a real option approach Federica Cucchiella and Massimo Gastaldi

700 Received August 2005 Revised January 2006 Accepted February 2006

Department of Electrical and Information Engineering, Faculty of Engineering, University of L’Aquila, L’Aquila, Italy Abstract Purpose – The aim of this paper is that of individualizing a framework for the management of uncertainty in supply chain finalized to reduce the firm risks. Design/methodology/approach – Since a way for reducing the damages deriving from uncertainty sources is increasing the level of flexibility inside the supply chain, and the real option theory allows the increase of the flexibility level, in order to achieve the aim of this work, we utilize the real options theory to coverage of one or more risks inside the supply chain. Findings – A useful theoretical framework has been individualized enabling the selection of possible options to protect the firm against the risk originating from every source of uncertainty. In particular, on two types of risks, using Matlab software, a test has been conducted that proves the ability of the outsource option to cover risks under examination. Practical implications – In the paper a framework providing useful information for the supply chain management is presented. Originality/value – The paper attempts to provide an original tool for the risks management deriving from production activities inside a supply chain. Keywords Supply chain management, Risk management, Supplier relations Paper type Research paper

1. Introduction In the actual economic environment, efficiency for manufacturing firms is moving from an internal to a supply chain priority. The relevance of right management of the supply chain is connected to the possibility that it offers to the firms to reach a competitive advantage on the market (Cousins and Spekman, 2003). Several studies on supply chain strategies (Handfield and Bechtel, 2004) and practice generally find that broader integration leads to improved performance, more specifically the investments made in supply chain have allowed to lean the productive process, to increase the level of consumers satisfaction, and to improve the inside productivity. Nowadays the competition is not among single firms, but on the contrary, among supply chains (Caputo et al., 2005; Sadler and Gough, 2005). According to several approaches in managerial literature, before selecting the correct strategy of supply chain, it is necessary, first of all, to understand which are the sources of uncertainty in the network, and in secondly, individualize the most correct way for reduce such level of uncertainty. At the same time, the individualization of the uncertainty sources is more and more complex due to the increasing complexity Journal of Manufacturing Technology Management Vol. 17 No. 6, 2006 pp. 700-720 q Emerald Group Publishing Limited 1741-038X DOI 10.1108/17410380610678756

The Authors would like to thank researches, organizers, chairmen at SCMIS 2005 and anonymous referees for their useful comments and suggestions. Moreover, authors want to inform that the paper has been supported by FIRB 2001 (Italian public funds for basic research), MIUR, Project Number: RBAU01L24Y_002.

revisable inside the supply chain network. This complexity is due to several sources: product/service complexity, e-business, outsourcing and globalisation (Harland et al., 2003). There are some factors concerning product/service that increase the supply chain complexity: scale, technological development, quantity of sub-systems components, degree of customisation of components in the final product/service, quantity of alternative design and delivery paths, number of feed back loops in the production and delivery systems, variety of distinct knowledge bases, skills and competencies incorporated in the product/service package, intensity and extent of the end-user involvement, uncertainty and change of end-user involvement and requirements, extend of supplier involvement in the innovation and transformation process, regulatory involvement, number of actors in the network, web of financial arrangements supporting the product/service, and extent of political and stakeholder intervention. Owing to the increasing product/service complexity it is possible that a firm cannot be excellent. E-business has emerged as a key enabler to realize the supply chain integration; through e-business the supply chain can gain global visibility across their extended network of trading partners and help them to respond quickly to market changes. By adopting e-business approaches firms can gain the benefits of supply chain integration – reduced costs, increased flexibility, faster response times – more rapidly and effectively, but at the same time, e-business introduces an higher level of complexity in the network connected to the substantial difficulty in realizing integrated processes and systems and providing rich information over distance. Considering another source of complexity, outsourcing, it is possible that the outsourcing can make necessary a change on supply network structure and processes. With reference to globalisation, some of the key supply chain challenges created by these sources of complexity include: greater constraints (longer lead times, minimum order quantities, etc.), greater volatility, uncertainty and risk, increased dependency on supply chain partners, reduced leverage, flexibility, responsiveness and control, less visibility across the supply chain, new cost structures, different set of competencies required to manage successfully. So, we can affirm that the sources of complexity inside a network are numerous and their management is more and more complex. At the same time, the high number of sources of complexity exposes the network to an increasing level of uncertainty and the uncertainty level exposes the network at an increasing number of risks. To achieve elevated levels of performance it is necessary to proceed to the correct management of such risks. This paper faces this problem, defining an approach, based on the real options theory, finalized to coverage the risks that arise from the sources of uncertainty existing inside the network. The paper is organized as follows: after this introduction on the increasing relevance of the supply chain management, Section 2 examines the sources of uncertainty inside the supply chain. In Section 3, after a bibliographic review of supply chain risks, from every source of uncertainty the associated risks are derived and links among them are presented. Real options are examined in Section 4. In Section 5 a useful framework individualizing the type of possible options able to protect the firm against the risks deriving from every source of uncertainty is presented. Finally, Section 6 closes the paper with a test, using Matlab software, on the ability of the individualized framework to cover the risks inside the supply chain.

Risk management in supply chain 701

JMTM 17,6

702

2. Sources of uncertainty inside the supply chain Several authors have paid attention on the sources of uncertainty, and the relative connected risks, inside a supply chain (Koh, 2004; Koh and Saad, 2004). In transaction costs theory uncertainty is connected to the variability of outcomes, lack of knowledge about the distribution of potential outcomes, and uncontrollability of outcome attainment. Likewise, in a supply chain there will be uncertainty every time that is not possible to define pre-emptive the ability of a supplier to realise the product with the characteristics required by the market, or, when the same supplier results incapable to quickly conform himself to the changes expected by the market. It is possible to identify three cluster of uncertainty sources inside a supply chain: order forecast horizon, input data and decision processes (Van der Vorst et al., 1998). The increasing complexity that characterizes the actual supply chains determines an always-increasing number of sources of uncertainty inside a network. The aim of the present paper is to individualize a useful framework for decrease the losses that arise from such sources of risk. Since, managerial literature individualizes an elevated number of sources of uncertainty that can be present in a network, it is necessary to define where to pay our attention. The selected sources of uncertainty that will be taken into account in the present work are: available capacity, manufacturing yield, supplier quality, internal organization, competitor action, information delays, stochastic cost, political environment, customs regulations and price fluctuations (Van Landeghem and Vanmaele, 2002). It is possible subdivide this set of sources of uncertainty in two subsets. The first one is referable to uncertainty sources internal at the network, whereas the second one, instead, is based on the external sources of uncertainty. The first subset is constituted by: available capacity, customs regulations, information delays and internal organization; the uncertainty sources of competitor action, manufacturing yield, political environment, price fluctuations, stochastic cost and supplier quality are in the second subset. Our research is organized in four steps as shown in Figure 1. Individualized a set of uncertainty sources, the analysis of the connected risks in the network is performed and finally the selection of real options able to protect the firm is presented. 3. Supply chain risks: a review The concept of risk has been extensively studied in literature from different perspectives. In Table I some risk definitions retrieved in literature are summarized with their bibliographical references. This table shows the wide attention that such topic has received from researchers and the following difficulty to contemporarily manage every risk. Now the sources of uncertainty presented in the previous section are taken back in examination and for each of them the associated risks inside the supply chain will be defined. Once such risks have been selected it is possible to hypothesize their coverage with one or more types of real options. So, now we examine the risks that every source of uncertainty presented in Table I can determine inside the supply chain and then individualize a framework for the coverage of the selected risks. One of the most important uncertainty sources is the available capacity. The absence of certainty related to the network availabilities can be examined on the basis of three different levels: financial, productive and structural. From a financial point of view, it is possible that the firm, due to an unavailability of financial resources, is unable to realize the product required by the market (Shapira, 1995). From a productive

Risk management in supply chain

Uncertainty

Internal sources uncertainty

External sources uncertainty

703 Competitor action Manufacturing yield Political environment Price fluctuations Stochastic cost Supplier quality

Available capacity Customs regulations Information delays Internal organization

Risk

Real options for risk coverage

point of view, the absence of certainty related to the available ability, can determine the risk that the project is too much extensive or too much complex with respect to the available network abilities. The risks related to the so-called bullwhip effects are an example of available capacity uncertainty. A further risk that the network run is structural and it is related to the unavailability of the infrastructures necessary to the realization and marketing of the product. The uncertainty sources connected to the internal organization on the lack of knowledge of the correct design of the supply chain, or, the not correct definition of the relationships among the actors embedded in the network. This can determine a low cooperation among the actors inside the network, or a low ability to adopt new technologies (Calabrese et al., 2005). Another source of uncertainty is related to the competitors actions; such actors, in fact, can develop a key role in the definition of the risks connected to the productive activity (Simons, 1999). The actions can, for instance, reduce and/or invalidate the competitive advantage that a firm holds, or, they can own particular skill that make difficult for the firm the realization of products that qualitatively are comparable with those realised by the competitors. From the absence of certainty around the manufacturing yields two types of risks can born. Firstly, it is possible that end consumer demand is lower than the business forecasts. Likewise, a too elevated consumers demand can equally result risky. In such case, in fact, if the firm is not able to adopt the necessary measures directed to extend the productive ability at the right moment, part of the consumer demand will be unsatisfied (Hallikas et al., 2004). Some risks, instead, can directly derive from the suppliers (Zsidisin et al., 2000). For

Figure 1. The steps of the research

Accounting ratios related to risk of ruin, default or bankruptcy Reduces utilisation of an asset and can arise when the ability of the asset to generate income is reduced Perceived risk differs if it is considered a new buy, modified rebuy, or straight rebuy The effects of factors such as age, professional organization membership, education, and job experience on risk perceptions Intrinsic motivational factors exist, such as the need for certainty, self-confidence, and the need to achieve, which affect individual risk perceptions The propensity to innovate, stability of the market structure, and growth rate affects risk perceptions The occurrence of performance risk is much higher among buyers in small companies Affects a firm’s ability to differentiate its products/services from its competitors Country of origin of buyer affects an individual’s risk preference Affects likelihood of customers placing orders, grouped with factors such as product obsolescence in “product market risk” The greater the risk involved, the greater the propensity to group, buy and share the risk involved The extent of communication or state of the relationship between a buyer and supplier influences the degree of perceived risk Risk being associated with negative outcome Exposes a firm to potential loss through changes in financial markets: can also occur when specific debtors defaults Arises through changes in taxation Risk perceptions differ according to the job and position of the buyer Exposes the firm to litigation with action arising from customers, suppliers, shareholders or employees

Accounting risk measures Asset impairment risk

Legal risk

Fiscal risk Job function

Downside risk Financial risk

Degree of customer/supplier interaction

Decision-making unit

Country Customer risk

Competitive risk

Characteristics of customer/supplier interaction Company size

Buyer’s personality

Buyer demographics

Meulbrook (2000)

Meulbrook (2000) Mitchell (1995)

Shapira (1995) Meulbrook (2000)

Mitchell (1995)

Mitchell (1995)

Mitchell (1995) Meulbrook (2000)

Simons (1999)

Mitchell (1995)

Mitchell (1995)

Mitchell (1995)

Mitchell (1995)

Mitchell (1995)

Baird and Thomas (1990) Simons (1999)

References

704

Buy type

Definition

Table I. Risks characteristics, definitions and references

Type of risk

(continued)

JMTM 17,6

The use of the capital asset pricing model to measure risk Affect a firm’s internal ability to produce and supply goods and services Risk taking is affected by the relationship between the company’s current position and some critical reference points Technical complexity and value of the item are positively correlated with the degree of perceived risk Exposes the firm with changes in regulations affecting the firm’s business, such environmental regulation Erodes value of whole business due to loss of confidence Risk cannot be captured with a single number, since multiple facets such as financial, technical, marketing, production and other risk aspect exist Strategies that could result in corporate disaster, bankruptcy or ruin Independence of action in venturing into the unknown Risk conditions equated with conditions characterized by newness, uncertainty, and lack of information Information scarcity as a key facet of uncertainty in terms of the existence of important resources and commitment duration Affect business strategy implementation Adversely affects inward flow of any type of resource to enable operations to take place The transpiration of significant and/or disappointing failures with inbound goods and services Adversely affects inward flow of any type of resource to enable operations to take place, also termed “input risk” Firm performance evaluated in terms of return and growth criteria Variability of the probability distribution of returns

Market risk Operations risk

Variance risk

Variability returns risk

Supply risks

Supply risk

Strategic risk Supply risk

Risk as lack of information

Risk as entrepreneurship Risk as innovation

Risk as disaster

Reputation risk Risk as a multi-faceted construct

Regulatory risk

Product characteristics

Organizational performance

Definition

Type of risk

Baird and Thomas (1990)

Baird and Thomas (1990)

Meulbrook (2000) and Smallman (1996)

Zsidisin et al. (2000)

Simons (1999) Harland et al. (2001)

Baird and Thomas (1990)

Baird and Thomas (1990) Baird and Thomas (1990)

Baird and Thomas (1990)

Meulbrook (2000), Bowen et al. (1998) and Smallman (1996) Schwartz and Gibb (1999) Shapira (1995)

Mitchell (1995)

Mitchell (1995)

Baird and Thomas (1990) Meulbrook (2000) and Simons (1999)

References

Risk management in supply chain 705

Table I.

JMTM 17,6

706

example, one risk is represented by the not detention, from the suppliers, of specific skills necessary for the realization of the product according to the characteristics required by the consumers. On the basis of risks classification inside a network, the technology is considered as one of the possible sources of the business risks. In fact, technological evolution is able to make obsolete the products that the firm already puts on the market. From a prices products point of view, some price fluctuations can be found when the firm is not able, both in the brief period that in the long period, to be profitable (Hallikas et al., 2002). The risks listed in Table II are those that we will be taken in examination in this work. In the next section the real options theory will be introduced in order to provide a tool for the coverage of considered risks inside a business reality. 4. Real option and supply chain Through the real options theory the management can firstly judge the strategies to adopt when inside the supply chain investments are made under uncertainty, and secondly make the right decision to give more flexibility to the production process. Real options theory presents a lot of similarities with financial options theory, and as the financial options also the real option can be split on the basis of their nature (Cohen and Huchzermeier, 1999) and gives the right but not the obligation, to practice a sale or a purchase. The success of the real options theory is connected to the increasing economic uncertainty environment in which firms operate. Uncertainty Internal sources Available capacity

Customs regulations Information delays

Internal Organization External sources Competitor action Manufacturing yield Political environment

Price fluctuations Table II. Risks deriving from uncertainty sources

Stochastic cost Supplier quality

The principals risks resulting from the selected uncertainty source Financial capacity: the project is not realizable of the excessive financial exposure Production capacity: the project is too much great or complex Structural capacity: the network does not have the necessary infrastructures Development from the consumers of a own product Not usability of product without right regulation Lack of information necessary for the right definition of product characteristics Lack of information necessary for the definition of the right moment of product emission on the market Not cooperation among the actors Low ability to adopt the new technologies Competitor actions can delete the achieved advantage Detention from competitors of a competitive advantage Low consumer demand of products Consumer demand of products superior than forecast An excessive demand of the consumer could make the mature product Not forecast of the possible actions of the vigilance authority Changes in the reference context can modify the type of demand products Not coverage of the costs sustained by the network due to product price fluctuations A new technology on the market could make obsolete the product Not availability of specific skills required to the suppliers

The introduction of the real options inside a supply chain allows the introduction of a greater level of flexibility in the network useful to cover the damages of market volatility. For instance, the damages from the uncertainty related to the level of product demand and the price/exchange rate, can be limited through the use of the real options (Cohen and Huchzermeier, 1999). Moreover, using a lattice-programming model, it is possible to quantify the benefits resulting from a greater degree of operational and managerial flexibility (Huchzermeier and Cohen, 1996). From literature emerges that: . the use of real options allows to improve the firm’s shareholder value and to reduce the level of risk that the firm races in the implementation of the production activities; and . real options need to be deployed, managed and exercised. It is now necessary to proceed to a description of some types of real options, in order to establish which type of real option suits better to every risks listed in Table II. Literature offers a wide number of real options types. In this paper only a subset will be take into account: defer, time or stage, explore, lease, outsource, alter operating state, abandon, growth, compound (Trigeorgis, 1996; Seppa¨ 2000). Owing to space limitation, in Table III the main characteristics of option are summarized with their biographical references. The following step is to associate every option to one or more risks recovered inside the supply chain (Benaroch, 2002). 5. Real option approach to limit firm risks In Section 3 the uncertainty risks inside the network have been individualized and in Section 4 some real options have been briefly described. At this point, a useful framework to limit risks in a network is shown in Figure 2, based on the following steps (Harland et al., 2003) (1) Analysis of supply chain. This step is finalized to an examination of the network structure, to define the most suitable performance measure and to delaine the responsibility inside the structure. (2) Identify uncertainty sources. After studying the supply chain structure, in this step it is possible to underline the most important sources of uncertainty that can cause losses for the firm. (3) Examine the subsequent risk. In this step we select the risks in the production activities. (4) Manage risk. At this point, with the aim to manage the risk, a preliminary analysis of risks inside the network and their damages is performed. (5) Individualize the most adequate real option. After the risks analysis, it is possible to select the real options more suitable able to coverage against the risks under examination. (6) Implement supply chain risk strategy. The last step is finalized to the implementation of the real option strategy defined in the previous steps. On the basis of the options characteristics and abilities, it is possible to individualize those able to provide coverage against one or more risks in the network (Table IV).

Risk management in supply chain 707

JMTM 17,6

708

Table III. Suggested real options for supply chain network management

Description

References

Defer: option to defer enables management to defer investment and benefit from more information. The management, with this kind of option, can wait x years to see if output prices justify constructing a building or a plant or developing a field Time or stage (stop-resume): when an investment can be see as a series of outlays, the stage option creates the opportunity to abandon the project in midstream if new information are unfavourable. Each stage can be viewed as an option on the value of subsequent stages and valued as compound option Explore (pilot/prototype): through an explore option is possible to realize a project on prototype scale. Both the costs that the payoff the prototype is proportional to those of the project realized on real scale. Observed the prototype results and on the base of the consequent results the management will decide if reply the project on real scale Lease: with this kind of option it is possible leasing or renting a property with an option to buy it at a future date. The future price of the property should be fixed at the time the lease-option is signed. Usually there is an up-front payment of some amount to purchase the option. The amount can vary. Sometimes the monthly payment is larger than normal and the excess is used to purchase the option Outsource: the resource required for the investments realization can be leased to external actors, in such way it is possible to transfer the risk of emergency costs or the costs due to an incapability to realize the investment in-house. Some times such option is connected to the stage option, in such way it is possible to avoid a penalty that would be imposed in the case of breakdown of the contract during the phase of realization Alter operating state: if market conditions are more favourable than expected, the firm can expand the scale of production or accelerate resource utilization. Conversely, if conditions are less favourable than expected, it can reduce the scale of operations. In extreme cases, production may be halted and restarted

Ingersoll and Ross (1992), Trigeorgis (1996) and Benaroch and Kauffman (2000)

Brennan and Swartz (1985), Trigeorgis and Mason (1987) and Pindyck (1988)

Clemons and Weber (1990), Kulatilaka et al. (1999)

Clemons and Weber (1990)

Richmond and Seidmann (1993)

Copeland et al. (1995), Trigeorgis (1996) and Kulatilaka et al. (1999)

(continued)

Description

References

Abandon (switch use): if market conditions decline severely, management can abandon current operations permanently and realize the resale value of capital equipment: it is an American put option on the project’s current value with an exercise price of salvage value Growth: an early investment (e.g. R&D, lease on undeveloped land or oil reserves, strategic acquisition, information network) is a prerequisite or a link in a chain of interrelated projects, opening up future growth opportunities (e.g. new product or process, oil reserves, access to new market, strengthening of core capabilities) Compound: real-life projects often involve a collection of various options. Upward-potential-enhancing and downward protection options are present in combination. Their combined value may differ from the sum of their separate values; i.e. they interact. They may also interact with financial flexibility options

Copeland et al. (1995), Trigeorgis (1996) and Hakan and Vandergraaf (1999)

Risk management in supply chain 709

Trigeorgis (1996), Zhu (1999) and Taudes et al. (2000)

Zhu (1999)

Table III.

Analyse supply chain • Structure of network • Key measures • Respondability inside SC

Identify uncertainty sources

Implement supply chain risk strategy

• Internal resources • External resources

Individualize the most adequate real option • • • •

Examine the subsequent risk • • • • •

Defer Stage Explore ……

Manage risk

• Risk position • Risk scenario

Likelihood of occurrence Stage inlife cycle Exposure Likely triggers Likely loss

Figure 2. Steps for risk management

Table IV. Mergers among real options and supply chain risks

Financial capacity: the project is not realizable of the excessive financial exposure Production capacity: the project is too much great or complex Structural capacity: the network does not have the necessary infrastructures Development from the consumers of a own product Not usability of product without right regulation Lack of information necessary for the right definition of product characteristics Lack of information necessary for the definition of the right moment of product emission on the market Not cooperation among the actors Low ability to adopt the new technologies

Internal organization External sources Competitor Competitor actions can delete the action achieved advantage Detention from competitors of a competitive advantage Manufacturing Low consumer demand of products yield

Information delays

Customs regulations

Internal sources Available capacity

The principals risks resulting from the selected uncertainty source

X

X

X

X X

X

X

X X

X

X

X

X

X X

X

X

X

X X

X

X X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

Real option types Scale Scale Defer Stage Explore Lease Outsource down up

X

X

X X

X

X

X

X

X

Abandon switch

710

Uncertainty sources

(continued)

Strategic grow

JMTM 17,6

Supplier quality

Price fluctuations Stochastic cost

Political environment

Uncertainty sources

Consumer demand of products superior than forecast An excessive demand of the consumer could make the mature product Not forecast of the possible actions of the vigilance authority Changes in the reference context can modify the type of demand products Not coverage of the costs sustained by the network due to product price fluctuations A new technology on the market could make obsolete the product Not availability of specific skills required to the suppliers

The principals risks resulting from the selected uncertainty source

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

Real option types Scale Scale Defer Stage Explore Lease Outsource down up

711

X

X

X

X

Strategic grow

X

X

X

Abandon switch

Risk management in supply chain

Table IV.

JMTM 17,6

712

For instance, considering the risk arising from an internal uncertainty source, that is the incapability of the network to satisfy the financial demands of the project, the management has four options that can limit the consequential losses from this kind of risk: defer, explore, lease, outsource. The defer option allows at the management to defer the moment in which begin an investment for a certain time period. In this case, the defer option is useful if the management is able in retrieving the necessary financial sources for the realization of the project. Alternatively, it possible to proceed through the use of an explore option; in fact, such option offers the possibility to test the profitability of the project on pilot/prototype scale; therefore, in absence of the financial resources required by the project in the integral scale, it can be useful to appraise the real project profitability. If the management evaluates that the payoff rising from the project are elevated, it can be proceed to an amplification of the financial resources. In the case in which the necessary financial resources are not enough and their amplification is not desired, it can be applied a lease option; in this case, in fact, the management of the network has to pay only the lease, while, all the purchase costs associated with the realization of the project will put out on the leased actor. Finally, the outsourced option allows to transfer outside the overrun risk, schedule overrun and the failure of an in-house realization of the project. When the firm faces the risk that the project is too great or complex, is possible to apply one of the following options: stage, explore, lease, outsourcing. The stage option allows dividing the project in a specific number of steps; therefore the production abilities required to the firm will be reduced. Otherwise, the explore option allows testing on prototype scale the profitability of the investment. Subsequently, on the basis of the results obtained by the test, it will be evaluated whether to proceed with the investment on real scale. The other two options, instead, are turned outside of the network. The first, lease option, allows the exploitation of the lease contracts for the amplification of production ability. The second, outsourcing option, foresees, instead, that a part of the project is realised by a third actor. Examining external sources of uncertainty (manufacturing yield), two risks for the firm can be derived: the demand of products from the consumers is lower than forecasts, or, the demand of the products is higher in comparison to the forecasts. In the first case, low demand of products, the firm can protect himself with a plurality of options. In fact, excluding the two options finalized to get the contraction of the productive scale, and the development of new investment opportunities, all the other options take into account can offer a protection against the risk of a too much low consumer demand. In the opposite case (high demand of products), three types of options are available: defer, explore e scale up. The first, permitting to defer the moment in which the project starts, allows to recover, in the deferral period, further information around the aim of purchase to the consumers. The explore option allows to make a test of the project on a pilot scale and therefore, to test the consumer demand of end products. Finally, with a scale up option is possible to proceed with an amplification of the production scale. Moreover, analysing Table IV it is possible to note that manager has at least two options for each of the 19 risks taken into account. In the two cases of “low consumer demand of products” and “not coverage of the costs sustained by the network due to product price fluctuations” it is possible to use seven types of options of the nine taken in examination with the purpose to reduce and/or to eliminate the damages derived from the presence of one of these two risks. Moreover, adopting a perspective analysis

based on the versatility of every option, the analysis highlights that the explore option is the most versatile option since it can furnish a protection against the damages derived from 17 risks of the 19 examined. The less utilized options are instead those of scale up and strategic growth. Such options, in fact, can usefully be utilized in only two cases of the 19 taken into account. Notice that in Table IV there are two risks underlined with an arrow in a black box; the capability of outsource option to cover these risks is tested in the next section. 6. Goodness test on real option approach This section is finalized to test the goodness of the proposed approach. More specifically, we want to prove the ability of the outsource option to cover, at the same time, the risks of production capacity and price fluctuations. To simulate such situation we considered a high technology company that produce medical devices (Sheffi, 2001). With the aim to make uncertain the economic environment in which such firm operates, the following hypothesis have been assumed: . The firm has a production plant that can be used to realise the products demand, but the production ability of such plant would be not able to cover, for a period time, the whole product demand (production capacity risk). . The product sale price could change, and in case of reduction, the new level price could be not enough for coverage the costs sustained for the products realization (price fluctuation risk). Under these hypotheses, the ability of outsource option to maximize the net present value (NPV) is tested. For such aim, through the use of software Matlab, the expected NPV is estimated for a firm characterized by the following assumptions: . The firm has a probability alpha (a) to be non able to satisfy, with the proper plant, the product demand. For each value of a – 0 there will be a not satisfied quote of demand. For example, for a ¼ 0.4 the firm probability to be able to satisfy the production demand is of 60 percent; to ensure the possibility to satisfy the demand, the firm can resort to an external supplier. . The product price is not defined, but it can change between 350 and 445e. With software Matlab the expected NPV gained by the firm has been estimated in three different cases: (1) total production demand realised with the proper plant; (2) total production demand realised through the external supplier plant; and (3) use of dual plants (proper and external) for the realization of production demand. Figure 3 shows the NPV of the firm if the manager decides to use the only proper plant for realize the whole production demand. With this production strategy, for a ¼ 1 the firm is not able to realize any production level and not having any external supplier, the NPV reaches the minimum value of zero. The maximum value of the NPV is obtained when the firm is able to realise all the production demand and the price of sale is equal to 445e.

Risk management in supply chain 713

JMTM 17,6

× 107 8 7

714 Net present Value

6 5 4 3 2 1

Figure 3. Expected net present value using the internal plant

450 430

410

0 0

0.2

390 0.4 α

0.6

0.8

370 1

Price

350

If the manager decides to use, for the whole products demand, the external supplier the NPV is not depending from a. As shown in Figure 4, the expected NPV is constant for every level of price. The last available production strategy foresees the joined use of own plant and external supplier plant (Figure 5).

× 107 8 7

Net present Value

6 5 4 3 2 1

Figure 4. Expected net present value using external supplier plant

0 0

450 410 0.2

390 0.4 α

0.6

0.8

370 1

350

Price

430

In this case the NPV, when probability is next to 1, can reach negative values. This is due to possible additional costs required by the external supplier to guarantee the production. At this point, for individualize the production strategy that maximize the NPV, in Figure 6 the three production strategies, have been jointly depicted.

Risk management in supply chain 715

× 107 8 7

Net present Value

6 5 4 3 2 1 0 –1 0

0.2

0.4 α

0.6

0.8

1

350

370

390

410

430

450

Figure 5. Expected net present value using dual plant

Price

× 107 8

Net present Value

6

4

2

0 0

450 0.2

430 0.4

390

0.6 α

0.8

370 1

350

410 Price

Figure 6. Expected net present value using selected production strategies

JMTM 17,6

716

Such figure suggests some decision strategies for the manager confirmed in Tables V and VI. Such tables points out, for every price and a levels the best production strategy. More specifically, if the product sale price is of 420e the optimal strategy, for the maximization of the NPV, it is the following: . proper plant for a , 0.10; . dual plant for 0.10 # a , 0.37; and . external plant for a $ 0.37. Moreover, it is possible to note that the greater costs sustained for the dual plant utilization are justified only beginning from a 370e price sale, and if a $ 0.37, it is useful directly turn to the external supplier without considering internal production. A similar analysis can be done observing Table VI. This table is more useful when the manager know the probability of proper plant to work. If, for example, there is an equal probability that the plant work or not work (a ¼ 0.5) is more proper the use of the external plant as soon as possible. In fact, in the case under examination, only for a price between 350 and 355e the extra cost of the external supplier determine a “NPV from external supplier” inferior than “NPV with proper plant”. But with a price higher than 355e the only implementing production strategy is related to the external plant.

Table V. Price based best production strategies

Table VI. Alpha based best production strategies

Price

Proper

350 360 370 380 390 400 410 420 430 440

a # 0.57 a , 0.43 a # 0.33 a # 0.23 a # 0.18 a , 0.14 a # 0.12 a , 0.10 a , 0.09 a , 0.08

a

Proper

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

350 # price , 425 350 , price , 385 350 , price , 375 350 , price , 365 350 , price , 355

Best production strategy. use of plant Proper and external

0.33 , a , 0.37 0.23 , a , 0.37 0.18 , a , 0.37 0.14 # a , 0.37 0.12 , a , 0.37 0.10 # a , 0.37 0.09 # a , 0.37 0.08 # a , 0.37

Best production strategy, use of plant Proper and external

External

a . 0.57 a $ 0.43 a $ 0.37 a $ 0.37 a $ 0.37 a $ 0.37 a $ 0.37 a $ 0.37 a $ 0.37 a $ 0.37

External plant

425 # price # 445 385 # price # 445 375 # price # 445 365 # price # 445 355 # price # 445 350 # price # 445 350 # price # 445 350 # price # 445 350 # price # 445 350 # price # 445

7. Final remarks In this paper, we have analyzed the application of a real option approach inside a supply chain operating in a uncertain environment. The relevance of this approach is already been underlined in literature since before the real option introduction the uncertainty was believed in industry as undesirable element for the strategic decision-making. By the use of real option firm management can deal with a different situation, indeed greater is the uncertainty, greater is the option value. The real option approach has recently covered the mainstream in the strategic management literature; this relevance is connected at the turbulent business environment in which the actual supply chains operate. Owing to this even increasing turbulence, the supply chain can gain better performance through an intense collaboration among the several actors, that is, several kinds of joints that coordinated effort between two parties in a supply chain for achieve a common goal are even more necessary. Effective collaboration within each entity and between entities in the supply chain is essential to achieve good performances. However, only a little number of collaboration initiatives have successful and the collaboration appears more and more difficult to achieve. A way for reduce the damages of collaboration failure is increase the supply chain flexibility and, with this regard, real option instrument can be used for reach such a scope. In so doing the management have the opportunity to adjust their decisions to the new situation that no one can foresee. However, real option application requires costs related to three different decision phases: (1) Discovery. In this phase, managers identify the areas where the most attractive opportunities of uncertainty are and which on them may potentially offer the greatest rewards from options. (2) Selection. In this phase, managers evaluate how to reach the possible means of providing flexibility to applications, and decide on which of these options should be implemented. (3) Monitoring. In the last phase, the managers should focus on the uncertainties that are weather in consistent with predictions or not. So that the decision makers will know when to implement or abandon the options that were built into system. Therefore, in terms of costs real option approach has disadvantages with respect to traditional discounted cash flow methodologies since, working on process flexibility, it requires more data on variability of considered parameters and models that well match the project under examination, it could be seen as a black box so not easily understood and utilized and finally, analysis is more complex and needs ad hoc computer programs to solve the real option algorithm. So, several problems can born working with real option and perhaps for these reasons applications are not numerous in literature (Lander and Pinches, 1998). Real option analysis has the most value when supply chain investment have marginal value (near 0 NPV), both uncertainty and project life are at least moderate and investment flexibility is present. In such case, without any doubt, this approach overcomes static and passive traditional analysis. In this paper, we have try to delaine a quantitative model useful for the evaluation of real option use, the described model is suitable for a specific supply chain that wish to cover chosen source of uncertainty. However, it can be modified and adapted for all type of uncertainty source.

Risk management in supply chain 717

JMTM 17,6

718

8. Conclusions In this paper we study a supply chain strategy for limiting the damages that can born from the sources of uncertainty recoverable inside a supply chain. For reach such purpose firstly a set of sources of uncertainty have been selected, subsequently the risks connected with each sources have been defined. At this point, with the aim to draw a framework for the risk management, we utilize the real option theory. It offers, in fact, to the managers some tools for the coverage of risks connected to the production activity. So, after the analysis of risk characteristics, it is possible to individualize the real options that better suit with the risk under consideration. Finally, using a programming software (Matlab), the outsource option has been tested to cover two risks related to production capacity and price fluctuations. References Baird, I.S. and Thomas, H. (1990), “What is risk anyway?”, in Bettis, R.A. and Thomas, H. (Eds), Risk, Strategy, and Management, JAI Press, Greenwich, CT, pp. 21-52. Benaroch, M. (2002), “Managing information technology investment risk: a real option perspective”, Journal of Management Information Systems, Vol. 19 No. 2, pp. 43-84. Benaroch, M. and Kauffman, R.J. (2000), “Justifying electronic banking network expansion using real options analysis”, MIS Quarterly, Vol. 24 No. 2, pp. 197-225. Bowen, F.E., Cousins, P.D. and Lamming, R.C. (1998), “The role of risk in environment related supplier initiatives”, paper presented at the 7th Annual IPSERA Conference, London. Brennan, M. and Swartz, E. (1985), “A new approach to evaluating natural resource investments”, Midland Corporate Finance Journal, Vol. 3 No. 1, pp. 37-47. Calabrese, A., Gastaldi, M. and Levialdi Ghiron, N. (2005), “Real option’s model to evaluate infrastructure flexibility: an application to photovoltaic technology”, International Journal of Technology Management, Vol. 29 Nos 1/2, pp. 173-91. Caputo, A.C., Cucchiella, F., Fratocchi, L. and Pelagagge, P.M. (2005), “An integrated framework for e-supply networks analysis”, Supply Chain Management: An International Journal, Vol. 10 No. 2, pp. 84-95. Clemons, E.K. and Weber, B. (1990), “Strategic information technology investments: guideline”, Journal of Management Information Systems, Vol. 7 No. 2, pp. 9-28. Cohen, M.A. and Huchzermeier, A. (1999), “Global supply chain management: a survey of research and applications”, in Tayur, S. and Ganeshan, R. (Eds), Quantitative Models for Supply Chain Management, Kluwer, Boston, MA, pp. 669-702. Copeland, T., Koller, T. and Murrin, J. (1995), Using Option Pricing Methods to Evaluate Flexibility, Wiley, New York, NY. Cousins, P.D. and Spekman, R. (2003), “Strategic supply and the management of inter- and intra-organisational relationships”, Journal of Purchasing & Supply Management, Vol. 9 No. 1, pp. 19-29. Hakan, E. and Vandergraaf, J. (1999), “Quantitative approaches for assessing the value of COTS-centric development”, paper presented at the 6th International Symposium on Software Metrics (METRICS’99), West Palm Beach, FL. Hallikas, J., Karvonen, I., Pulkkinen, U., Virolainen, V.M. and Tuominen, M. (2004), “Risk management processes in supplier networks”, International Journal of Production Economics, Vol. 90, pp. 47-58.

Hallikas, J., Virolainen, V.M. and Tuominen, M. (2002), “Risk analysis and assessment in network environments: a dyadic case study”, International Journal of Production Economics, Vol. 78, pp. 45-55. Handfield, R.B. and Bechtel, C. (2004), “Trust, power, dependence, and economics: can SCM research borrow paradigms?”, International Journal of Integrated Supply Management, Vol. 1 No. 1, pp. 3-32. Harland, C., Brenchley, R. and Walker, H. (2003), “Risk in supply networks”, Journal of Purchasing & Supply Management, Vol. 9 No. 1, pp. 51-62. Harland, C.M., Knight, L.A. and Sutton, R.Y. (2001), “Information for supply interventions: sector, network and organization opportunities from network and organization opportunities from e-commerce”, paper presented at the 10th International Annual Conference of International Purchasing and Supply Education and Research Association, Jo¨nko¨ping, Sweden. Huchzermeier, A. and Cohen, M. (1996), “Vauling operational flexibility under exchange rate risk”, Operations Research, Vol. 44 No. 1, pp. 100-13. Ingersoll, J. and Ross, S. (1992), “Waiting to invest: investment and uncertainty”, Journal of Business, Vol. 65 No. 1, pp. 1-29. Koh, S.C.L. (2004), “MRP-controlled batch-manufacturing environment under uncertainty”, Journal of The Operational Research Society, Vol. 55 No. 3, pp. 219-32. Koh, S.C.L. and Saad, S.M. (2004), “Modelling uncertainty under a multi-echelon ERP-controlled manufacturing system”, International Journal of Integrated Manufacturing Systems, Vol. 15 No. 3, pp. 239-53. Kulatilaka, N., Balasubramanian, P. and Strock, J. (1999), “Using real options to frame the IT investment problem”, in Trigeorgis, L. (Ed.), Real Options and Business Strategy: Applications to Decision-Making, England Risk Books, London. Lander, D.M. and Pinches, G.E. (1998), “Challanges to the practical implementation of modeling and valuing real option”, The Quarterly Review of Economics and Finance, Vol. 44 No. 5, pp. 751-67. Meulbrook, L. (2000), “Total strategies for company-wide risk control”, Financial Times, Vol. 9. Mitchell, V.W. (1995), “Organizational risk perception and reduction – a literature review”, British Journal of Management, Vol. 6, pp. 115-33. Pindyck, R. (1988), “Irreversible investment, capacity choice and the value of the firm”, The American Economic Review, Vol. 78 No. 5, pp. 969-85. Richmond, W.B. and Seidmann, A. (1993), “Software development outsourcing contract structure and business value”, Journal of Management Information Systems, Vol. 10 No. 1, pp. 57-72. Sadler, I. and Gough, R. (2005), “Applying a strategic planning process to several supply chain partners”, Journal of Manufacturing Technology Management, Vol. 16 No. 8, pp. 890-908. Schwartz, P. and Gibb, B. (1999), When Good Companies do Bad Things, Wiley, New York, NY. Seppa¨, T. (2000), An Option-Pricing Approach to Risk and Return in Venture Capital, Hseba, Helsinki. Shapira, Z. (1995), Risk Taking: A Managerial Perspective, Russell Sage Foundation, New York, NY. Sheffi, Y. (2001), “Supply chain management under the threat of international terrorism”, International Journal of Logistics Management, Vol. 12 No. 2, pp. 1-11. Simons, R.L. (1999), “How risky is your company?”, Harvard Business Review, Vol. 77 No. 3, pp. 85-95.

Risk management in supply chain 719

JMTM 17,6

720

Smallman, C. (1996), “Risk and organisational behaviour: a research model”, Disaster Prevention and Management, Vol. 5 No. 2, pp. 12-26. Taudes, A., Feurstein, M. and Mild, A. (2000), “Options analysis of software platform decisions: a case study”, MIS Quarterly, Vol. 24 No. 2, pp. 227-43. Trigeorgis, L. (1996), Real Options: Managerial Flexibility and Strategy in Resource Allocation, MIT Press, Cambridge, MA. Trigeorgis, L. and Mason, S.P. (1987), “Valuing managerial flexibility”, Midland Corporate Finance Journal, Vol. 5 No. 1, pp. 14-21. Van der Vorst, J., Beulens, A., De Wit, W. and Van Beek, P. (1998), “Supply chain management in food chains: improving performance by reducing uncertainty”, International Transactions in Operational Research, Vol. 5 No. 6, pp. 487-99. Van Landeghem, H. and Vanmaele, H. (2002), “Robust planning: a new paradigm for demand chain planning”, Journal of operations Management, Vol. 20, pp. 769-83. Zhu, K. (1999), “Evaluating information technology investment: cash flows or growth options?”, paper presented at Workshop on Information Systems Economics (WISE’99), Charlotte, NC. Zsidisin, G.A., Panelli, A. and Upton, R. (2000), “Purchasing organization involvement in risk assessments, contingency plans, and risk management: an exploratory study”, Supply Chain Management: An International Journal, Vol. 5 No. 4, pp. 187-97. About the authors Federica Cucchiella graduated from the Faculty of Economics and Business in L’Aquila (Italy). In 2004, she finalized her studies with a PhD in Managerial Engineering at the Department of Economics and Technology in Republic of San Marino University. She is now post PhD student at Department of Electrical and Information Engineering, University of L’Aquila. Her principal area of research regards supply chain management and real option. Federica Cucchiella is the corresponding author and can be contacted at: [email protected]. Massimo Gastaldi is Associated Professor in industrial economics at the Department of Electrical and Information Engineering, University of L’Aquila, where he also teaches industrial economics, analysis of financial systems and service economy. Since, 1988, he has contributed to several publications and edited books working for National Research Council (C.N.R.) and at University of Rome “Tor Vergata”. His principal area of research regards networks regulation, public utilities, supply chain and real options. He is responsible manager of private and public research projects. Massimo Gastaldi can be contacted at: [email protected].

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

The current issue and full text archive of this journal is available at www.emeraldinsight.com/1741-038X.htm

Managing stock-outs effectively with order fulfilment systems

Managing stockouts effectively

Richard Pibernik MIT-Zaragoza International Logistics Program, Zaragoza Logistics Center, Zaragoza, Spain Abstract

721 Received August 2005 Revised January 2006 Accepted March 2006

Purpose – When approaching a stock-out situation, a company should be able to actively manage the allocation of available products on the basis of customer requirements and priorities as well as contractual relationships. The purpose of this paper is to describe different order promising mechanisms and analyze how well they can contribute to the effective management of stock-out situations. Design/methodology/approach – The paper provides a formal description and analysis of alternative order promising mechanism applicable in make to stock systems. Numerical analysis is conducted based on the data of a pharmaceutical company. Findings – The paper clearly points out the potential of alternative order promising mechanisms to alleviate the negative consequences associated with a temporary stock-out situation. Research limitations/implications – The paper does not consider implications of inventory pre-allocation to customer classes. Further research should address the interplay between pre-allocation and different order allocation mechanisms. Practical implications – The results obtained from this analysis provide guidelines for manufacturers, retailers, and vendors of supply chain software on how to design and utilize order promising systems. Originality/value – The paper provides a consistent formal approach to modelling order promising mechanisms, introduce new and innovative order promising mechanisms and provide valuable insight into their performance through numerical analysis. Keywords Order processing, Stock control, Resource allocation, Customer requirements Paper type Research paper

Introduction In recent years, companies have been striving for increased supply chain efficiency through higher resource utilization and inventory reduction. To a certain extent, efficient supply chains with limited slack and buffer in terms of capacity and inventory become more vulnerable in regard to demand fluctuation, supply shortages and uncertain manufacturing yields. In this context, the management of stock-out situations becomes an important issue. A company should be able to anticipate stock-out situations and should actively manage the allocation and re-allocation of available products on the basis of customer requirements and priorities as well as contractual relationships. Order fulfilment systems, which are increasingly being employed, play an important role in managing stock-out situations. These commercial applications provide functions for receiving orders through multiple channels (e.g. call centers, e-mail, internet platforms), for order promising, i.e. for checking product availability and quoting delivery quantities and due dates, as well as for order execution and monitoring. When approaching a stock-out situation, order promising as part of the overall order fulfilment

Journal of Manufacturing Technology Management Vol. 17 No. 6, 2006 pp. 721-736 q Emerald Group Publishing Limited 1741-038X DOI 10.1108/17410380610678765

JMTM 17,6

722

process is crucial. It determines how the remaining available inventory/capacity is allocated to incoming orders. These allocation decisions can have significant impact on a company’s short and long term profitability. In the short run, not meeting customer requirements in terms of delivery dates and delivery quantities can, for example, cause lost sales, and contractual penalties. Long-term impacts can be decreased customer retention rates and lower future sales potential, resulting in diminishing customer lifetime values. In a stock-out situation, the order promising logic has to take these shortand long-term implications into account; order promising decisions should be made in such a way that overall negative impacts on profitability are minimized. In this paper, we describe different order promising mechanisms and analyze how well they can contribute to effectively managing a stock-out situation. We characterize the logic implemented in commercial order fulfilment systems and identify and evaluate alternative allocation mechanisms. One basic idea we are pursuing is that a company may be able to switch from a standard mode of order promising into an “emergency mode” as soon as a stock-out situation is anticipated. We utilize data from a specific stock-out situation in the pharmaceuticals sector to conduct numerical analysis. The results obtained will illustrate the effects of different allocation mechanisms and provide insight into their short- and long-term implications. Literature review So far, only a limited number of publications address order fulfilment in stock-out situations. Traditional inventory management literature mainly focuses on optimal stocking decisions when facing uncertain demand. Stock-out costs are modeled as cost of lost sales or backorder cost on a per unit (and time) or per stock-out occasion basis. Alternatively, they are incorporated implicitly through a service level constraint. A number of publications (Ha, 1997; de Ve´ricourt et al., 2002) propose approaches to pre-allocation of inventory to different customer classes (so-called inventory rationing). By making pre-allocation decisions based on individual stock-out cost, the authors acknowledge that cost resulting from unfilled demand can depend on specific types of customers, and that available inventory should be allocated in such a way that overall cost are minimized. In contrast to pre-allocation, where demand of different customer classes is considered to be uncertain, order fulfilment has to allocate received orders to available inventory/capacity in order to determine order specific due dates and delivery quantities. Major contributions in this field of research have been made by Ball et al. (2004), Chen et al. (2002) and Chen et al. (2001). Next to an overview of so-called “available to promise” functions in commercial order fulfilment systems and of a number of industry applications, the authors develop mixed-integer programming models for allocating orders arriving within a pre-determined time interval to available components and assembly capacity. The proposed models are specifically addressing order promising in a configure to order environment. Allocation decisions are made on the basis of “profits” calculated by subtracting tangible and intangible cost, which also account for penalties induced by order denial, from overall revenue. Fleischmann and Meyr (2003) develop linear and mixed integer programming models for allocating orders to finished goods inventory as well as available resources of an assemble to order system. Allocation decisions are based on a penalty cost parameter that includes costs for backlogging, early allocation and order denial. Kilger and Schneeweiss (2000) give an overview of the order allocation functions provided by commercial order

fulfilment systems. They point out that order allocation is typically performed on the basis of a set of heuristic allocation rules that can be customized to accommodate company specific requirements. Due-date setting in a make to order environment has been addressed in numerous publications (an overview is provided by Gordon et al., 2002). As this paper focuses on the allocation of finished goods inventory, we will omit a further discussion of the proposed approaches. Although, some techniques for allocating orders to available inventory have been proposed, their performance has not been assessed in regard to stock-out situations. The research presented in this paper specifically focuses on order allocation in a stock-out situation. We analyze how well different allocation mechanisms can contribute to effectively managing a stock-out situation. In our analysis, we include both heuristic allocation mechanisms, implemented in commercial order fulfilment systems and an optimization-based approach similar to the models proposed by Chen et al. (2001, 2002). Order fulfilment in stock-out situations One important task of an order fulfilment system is the allocation of incoming orders to available inventory/capacity while accounting for customer preferences in regard to delivery quantities and due dates. In the following, we will focus on a manufacturer/retailer delivering goods from a single production facility or warehouse. For simplicity, we will only consider one product which is being produced to stock. Finished goods inventory supply is assumed to be fixed in the short term, determined by planned receipts from production or purchase. For modeling the allocation problem, we assume a discrete time structure with a planning horizon spanning over T þ 1 periods. With tb, we denote the current period in which orders are being received and allocation decisions have to be made. With t e ¼ tb þ T; we denote the final period of the planning horizon. Typically, the periods t [ [tb,te] represent working days for which due dates of individual orders are being quoted. In the current period tb, the order fulfilment system receives a number of orders. With O (tb), we will denote the set of orders received in period tb. We assume that each order i [ O (tb) can be characterized by a triple ðqi ; dli ; d ui Þ of ordered quantity qi, earliest date of delivery dli and latest date of delivery dui ; with dli ; d ui [ ½tb þ 1; te
specified by the customer. In period tb, the planned inventory invt ðtb Þ available in t can be calculated on the basis of the initial inventory invtb ðt b Þ at the end of the current period, the planned receipts st, determined by MPS or procurement orders, and inventory quantities rt, which have been reserved for orders committed previously: invt ðt b Þ ¼ invtb ðt b Þ þ

t X

st 2

t¼t b þ1

t X

rt


 ;t [ t b þ 1; t e

ð1Þ

t¼t b þ1

The “available to promise” quantity can be calculated as atpte ðt b Þ ¼ invte ðtb Þ for period te and atpt ðt b Þ ¼ minfinvt ðt b Þ; atptþ1 ðtb Þg for periods te21 ; . . . ; t bþ1 : Defining by xi (t) the quantity quoted for order i [ Oðt b Þ in period t, we can represent the result of the allocation of all orders received in period tb by a (jO(tb)j £ T 2 1)-dimensional matrix xtb : xtb is a feasible allocation if the condition: invtb ðtb Þ þ

t X

t¼t b þ1

st 2

t X

t¼t b þ1

rt 2

t X X

i[Oðt b Þ t¼t b þ1

xi ðtÞ $ 0

ð2Þ

Managing stockouts effectively

723

JMTM 17,6

724

is satisfied for all t [ [tb þ 1,te]. Order allocation is typically performed on an infinite rolling time horizon basis. In every period, a set of orders is allocated to the available to promise inventory. The individual allocations are linked through the available to promise quantities: The allocation xt0b determines the available to promise quantities atpt ðt00b Þ (;t [ ½t 00b þ 1; t00e
) for the allocation xt00b in period t 00b ð¼ t 0b þ 1Þ; etc. From an order fulfilment perspective, a stock-out occurring in a specific period t 0b can be defined as a situation, in which atpt ðt 0b Þ is zero (or less than a minimum level) for a time period ½_t ;
t
[ ½tb þ 1; te
(stock-out period). As a consequence, incoming orders i [ Oðt 0b Þ with a required delivery time window bd li ; dui c [ ½t b ;
t
cannot be fulfiled according to customer requirements. If no prior measures are taken, the remaining available to promise inventory atpt ðtb Þ for periods t [ ½_t ;
t
is consumed by orders received in periods . . . ; t0b 2 2; t0b 2 1 until a specific period tb0 is reached, in which at least one order cannot be fulfiled. The number of subsequent periods t 0b þ 1; t 0b þ 2; . . . in which all or a subset of orders of O (tb0 þ 1), O (tb0 þ 2), . . . remain unfilled depends on the overall order quantities in these periods as well as the dates and quantities of planned receipts, which determine the duration of the stock-out period ½_t;
t
: When approaching a stock-out situation, two questions are of major importance: (1) How should the remaining quantities atpt ðt0b Þ available in ½_t ;
t
be allocated to orders i [ Oðt 0b Þ if not all orders can be fulfiled according to customer requirements? (2) Which appropriate measures can be taken prior to the stock-out situation, enabling a company to effectively allocate orders i [ Oðt 0b þ 1Þ, Oðt 0b þ 2Þ to remaining available to promise quantities? Whereas the first question mainly refers to mechanisms for generating an allocation xt0b ; the second question relates to approaches which shift consumption of the available to promise inventory to later periods. If remaining available to promise inventory is fixed (i.e. planned receipts are given and orders committed previously are not rescheduled), the main task of an order promising system is the allocation of orders to the given remaining inventory quantities in such a way that the overall negative consequences for a company are minimized. An order fulfilment system should therefore be able make order promising decisions on the basis of order specific effects on a company’s short- and long-term profitability. In contrast to planning tasks, in which decisions are commonly based on aggregate measures (e.g. demand fill rates used in inventory management), order-promising decisions should be based on order specific measures, capturing the short- and long-term effects of not meeting customer requirements. Short-term consequences can be lost profits, contractual penalties and additional cost, e.g. for expediting or upgrading orders. Generally, these short-term implications can easily be captured in a sufficiently accurate way. Long-term implications are more difficult to account for. In the long run, late fulfilment can result in lower future sales potential and a decreased retention rate. Customer dissatisfaction and consequently also the reaction in terms of future order placement are specific to individual customers and will especially depend on the tardiness of the quoted due date. If we assume that order specific short- and long-term consequences csi ðxi ðt b þ 1Þ; . . . ; xi ðt e ÞÞ and cli ðxi ðtb þ 1Þ; . . . ; xi ðte ÞÞ can be quantified, a value Cðxtb Þ reflecting overall negative consequences can be assigned to any allocation xtb : The objective of an order fulfilment system is to

generate a sequence of allocations . . . ; xt0b 21 ; xt0b ; xt0b þ1 ; . . . minimizing the total negative consequences for the time, a stock-out situation persists. The most precise approach to capturing cli ð · Þ would be based on an estimate of the impact, order promising decisions have on customer lifetime value. The decrease in customer lifetime value induced by assigning a specific (late) due date would then have to be estimated for every individual order. In most cases, however, this approach will not be practicable. Apart from the fact that only few companies actually use customer lifetime value approaches, it is unlikely that a mapping between the decrease in customer lifetime value and order specific tardiness or partial fulfilment measures can be established. In the following section, we will analyze alternative allocation mechanisms and examine how well order promising decisions are aligned with the objective of minimizing overall negative impacts. Order promising logic FCFS allocation In order fulfilment systems, “first-come-first-served” (FCFS) is the predominant form of allocating orders to available to promise inventory. The incoming orders are allocated in the sequence of their arrival; the allocation of a specific order i0 [ Oðtb Þ to available to promise inventory atpt ðt b Þ is performed immediately after the order arrives. After every allocation of an order, atpt ðt b Þ is recalculated. If we assume that the set Oðtb Þ is ordered and that index i reflects the sequence in which orders arrived, the basic FCFS allocation mechanism for any specific order i0 (i.e. the i0 -th order received in period tb) can be described as follows: (1) Initialization Customer order i0 has been received. Record customer requirements ðqui0 ; d ui0 ; dli0 Þ: (2) Check availability Determine atpt ðtb Þ for all t [ ½t b þ 1; t e
: Is atpt ðt b Þ $ qi0 for at least one t [ ½d li0 ; d ui0
? yes: go to 3a. no: Is atpt ðt b Þ $ qi0 for at least one t [ ½d li0 þ 1; te
? yes: order late; go to 3b. no: reject order; end. (3) Assign due date d^ i0 n h io 3a: xi0 ðd^ i0 Þ ¼ qi0 with d^ i0 ¼ min t jatpt ðt b Þ $ qi0 ; t [ dl 0 ; du0 i

3b: xi0 ðd^ i0 Þ ¼ qi0

with

n

h

i

d^ i0 ¼ min tjatpt ðt b Þ $ qi0 ; t [ dli0 þ 1; t e

io

:

Managing stockouts effectively

725

JMTM 17,6

726

(4) Reservation and confirmation. Reserve quantity xi0 ðd^ i0 Þ: Confirm due date. Update rt. End. Due to the fact that the FCFS allocation will commit to every incoming order as long as inventory is available to promise, it cannot account for order specific short- and long-term consequences. When approaching a stock-out in ½_t ;
t
, an order promising system employing FCFS will consume atpt ðt0b Þ ðt [ ½tbþ1 ;
t
Þ until it reaches a point after which all subsequent orders i [ Oðt 0b Þ with bdli ; dui c [ ½t bþ1 ;
t
will be quoted late or will have to be rejected. FCFS is an appropriate allocation mechanism as long as available inventory is sufficient to fulfil all incoming orders. In a stock-out situation, however, FCFS allocation is not aligned to the objective of minimizing overall negative consequences. Commercial order fulfilment systems offer partial delivery generation. If the quantity qi0 of an order i0 is not available in bdli0 ; dui0 c, the order can be split up and fulfiled through two or more partial deliveries. Generally, the due date of the first partial delivery should be quoted within the time interval bdli0 ; d ui0 c and be at least of size a · qi0 ; where að0 , a , 1Þ defines the minimum fraction of the order quantity to be fulfiled with the first partial delivery. The FCFS allocation mechanism can easily be extended to account for partial deliveries: 3b. Partial deliveries Is atpt ðtb Þ $ a · qi0 for at least one t [ bdli0 ; d ui0 c? 1 1 yes: xi0 ðd^ i0 Þ ¼ atpt ðt b Þ with t ¼ d^ i0 ¼ d ui0 . Determine atpt ðt b Þ for all t [ bdui0 þ 1; te c. Is atpt ðtb Þ $ qi0 2 xi0 ðd^ i0 Þ for at least one t [ bdli0 þ 1; t e c? 2 1 2 1 yes: xi0 ðd^ i0 Þ ¼ qi0 2 xi0 ðd^ i0 Þ with d^ i0 ¼ min {tjatpt ðt b Þ $ qi0 2 xi0 ðd^ i Þ}. no: xi0 ðdui0 Þ ¼ 0; xi0 ðd^ i0 Þ ¼ qi0 with d^ i0 ¼ min{tjatpt ðtb Þ $ qi0 ; t [ bdli0 þ 1; t e c}. no: go to 4. Incorporated into an FCFS allocation mechanism, the potential for generating partial deliveries will be limited. Depending on the minimum delivery quantity a · qi partial deliveries will only be applicable to very few orders. Prior to a stock-out period ½_t;
t
, FCFS allocation will fulfil every incoming order i with bd li ; dui c [ ½tbþ1 ;
t
as long as atpt ðtb Þ $ qi for at least one t [ bdli ; dui c. For the first order i0 in the sequence of incoming orders with qi0 . atpt ðtb Þ;t [ bd li ; dui c and atpt ðtb Þ $1 qi · a for at least one t [ bd li ; dui c; the system will generate a partial delivery xi0 ðd^ i0 ¼ dui0 Þ $a · qi0 if the 1 ^ remaining quantity qi 2 xi0 ðdi0 Þ can be quoted for a period t[ ½
t þ 1; t e
. After 1 consuming xi0 ðd^ i0 Þ; the ability of the system to generate further partial deliveries depends on the order quantities qi 0 of subsequent orders i00 . i0 with required delivery dates bd li00 ; dui00 c[ ½tbþ1 ;
t
. One way to attenuate the effects of a stock-out situation is to enforce partial deliveries. As soon as the stock-out situation in ½_t;
t
is anticipated, the order fulfilment

system can be operated in a mode, in which partial deliveries are automatically generated either for all or for certain pre-defined orders. Consumption of available to promise inventory will then be shifted to later periods t [ ½
t þ 1; te
: Rank-based allocation Next to FCFS allocation, order fulfilment systems commonly support a rank-based approach. When employing a rank-based allocation mechanism, incoming orders are first collected and then sequentially allocated to the available to promise inventory. The sequence in which orders are allocated is defined by a ranking, which is typically determined on the basis of a parameter reflecting the (relative) priority of the customer placing the order. From an order fulfilment system’s viewpoint, customer priorities are usually endogenous, provided by CRM or sales applications. Calculated on the basis of historical sales volumes or margins, they are originally determined to support allocation of sales and marketing resources to individual customers or customer classes. For the individual orders, the allocation mechanism is identical to the FCFS mechanism described previously; only the sequence in which orders are allocated is being altered, i.e. the order book is re-sorted based on customer priorities. Therefore, we omit a formal description of the allocation mechanism. In comparison to FCFS allocation, a rank-based allocation mechanism will increase the average response time to customer requests. Thus, rank-based allocation does not support real time order promising. Facing a stock-out situation, however, a company may choose to switch from a real time FCFS allocation to an “emergency batch mode” if this contributes to a more beneficial allocation of orders to available to promise inventory. At the cost of an increased average response time, a rank-based allocation mechanism allows for a prioritization of customer orders. A higher ranking will increase the likelihood of the order being fulfiled according to customer requirements. The effectiveness of rank-based allocation depends on how well the priorities assigned to the orders represent the consequences of not meeting customer requirements. Assuming that customer priorities perfectly reflect the negative consequences, rank-based allocation appears to be an appropriate approach to allocating customer orders. However, customer priorities are typically not determined to support allocation of customer orders to available to promise inventory but rather to support marketing decision. Even if the priorities do reflect future sales potential as one component of customer lifetime value, they may not adequately capture possible customer reactions to late or incomplete fulfilment. It is a moot question whether customer priorities obtained from CRM systems and the impact of order promising on customers’ retention rate are strongly related. Another inherent problem of a rank-based approach is prevalent when order quantities vary significantly. The sequential allocation does not take into account that due setting for one individual order has an impact on the system’s capability to fulfil remaining orders with a lower rank. As a consequence, a large order may be committed at the cost of having to delay or reject several smaller orders of the same or a lower priority class. Not taking these interdependencies between individual order promising decisions into account may lead to an allocation of orders to available to promise inventory, which is not aligned with the objective of minimizing overall negative effects induced by a stock-out situation.

Managing stockouts effectively

727

JMTM 17,6

728

Optimization-based allocation In this section, we introduce an optimization-based approach for allocating orders collected in period tb to available to promise inventory. In comparison to a rank-based allocation, an optimization-based approach allows for more detailed modeling of the short- and long-term effects of order promising. By simultaneously allocating the orders to the available to promise inventory it considers the interdependencies between order promising decisions for individual orders and can therefore alleviate inherent problems of an FCFS and rank-based approach. For allocating orders to available to promise inventory, we utilize a mixed integer programming model which allows for order rejection, late and partial deliveries. For partial deliveries, we make the same assumptions as previously, i.e. a maximum of two partial deliveries, first delivery is quoted for a period t [ bdli ; dui c and a minimum quantity of a · qi for the first delivery. In addition to the notation used previously, we introduce the binary variables u1i ðtÞ and u2i ðtÞ to specify due dates for the first and second delivery. u1i ðtÞ takes on the value 1 if the due date of the first (partial) delivery of order i is t and 0 otherwise. Likewise, u2i ðtÞ is defined for the second partial delivery. With binary variable vi, we model acceptance/rejection decisions. vi indicates whether a due date has been assigned to order iðvi ¼ 1Þ or not (vi ¼ 0). With Ddi, we denote the tardiness of an order as the positive deviation of the quoted due date for the first delivery (denoted by d^ i ) from dui . The following mixed-integer-programming formulation can be employed for determining due dates and order quantities for a set of orders Oðt b Þ : X X csi ðx1i ; x2i Þ þ cli ðx1i ; x2i Þ min C ¼ ð3Þ i[Oðt b Þ

i[Oðt b Þ

s.t. x1i ðtÞ $ a · qi · u1i ðtÞ

;i [ Oðtb Þ; t [ bd li ; dui c

x1i ðtÞ # qi · u1i ðtÞ ;i [ Oðtb Þ; t [ x1i ðtÞ

¼

qi · u1i ðtÞ

;i [ Oðt b Þ; t [

ð4Þ

bdli ; dui c

bdui

ð5Þ

þ 1; t e c

x2i ðtÞ # ð1 2 aÞ · qi · u2i ðtÞ ;i [ Oðt b Þ; t [ bt b þ 1; t e c te X

ðx1i ðtÞ þ x2i ðtÞÞ ¼ qi · vi

ð6Þ ð7Þ

;i [ Oðt b Þ

ð8Þ

 x1i ðtÞ þ x2i ðtÞ $ 0


 ;t [ t b þ 1; t e ð9Þ

t¼dli

invtb þ

t X

t¼t b þ1

st 2

t X

t¼t b þ1

rt 2

X

t X 

i[Oðt b Þ t¼t b þ1 te X

u1i ðtÞ ¼ vi

;i [ Oðt b Þ

ð10Þ

t¼d li te X t¼dui þ1

u2i ðtÞ # 1

;i [ Oðtb Þ

ð11Þ

d^ i ¼

te X

u1i ðtÞ · t

;i [ Oðtb Þ

ð12Þ

t¼dli

Ddi ¼ maxð0; d^ i 2 d ui Þ ;i [ Oðt b Þ
 u1i ðtÞ; u2i ðtÞ [ f0; 1g ;i [ Oðtb Þ; t [ tb þ 1; te vi [ f0; 1g

;i [ Oðt b Þ

ð13Þ ð14Þ ð15Þ

The generic objective function (3) accounts for both for short-and long-term effects of order promising. The model’s objective is to determine an allocation for orders i [ O(tb) that overall negative effects are minimized. Below we will propose an adequate formulation representing short- and long-term consequences. Constraints (4) and (5) define lower and upper bounds for the first (partial) delivery. Constraints (6) ensure a delivery quantity of qi for all orders fulfiled late, i.e. no late (first) partial deliveries are quoted. The logical constraints (7) link the variables u2i ðtÞ and x2i ðtÞ that specify the second partial deliveries quoted for a period t [ bdui þ 1; t e c: Constraints (8) ensure that quantity qi is assigned to every accepted order. Constraints (9) ensure non-negativity of planned inventory. Constraints (10) link due date quoting decisions with order acceptance/rejection decisions. Constraints (11) ensure that only one second partial delivery is quoted. Constraints (12) and (13) provide quoted due dates and tardiness. Equations (14) and (15) are integrality constraints for the binary variables. As pointed out previously, order allocation is performed on a rolling time horizon basis. In any period t 0b ; the model is employed to allocate orders Oðt 0b Þ to the available to promise inventory. Thereupon, atpt ðt 00b ¼ t0b þ 1Þ is determined for all t [ ½t 00b þ 1; t00e
and utilized for allocating the set of orders Oðt00b Þ arriving in t00b : This procedure is repeated in all following periods in which orders are received. The allocation of orders to available to promise inventory is dependent on the specific definition of the terms representing short- and long-term effects of order promising. One possible representation of the short-term consequences can, for example, be: u

csi ðx1i ; x2i Þ

¼ mi · pðDdi Þ þ cpi · ð1 2

di X t¼d li

u1i ðtÞÞ þ

te X

tc · u2i ðtÞ

Managing stockouts effectively

ð16Þ

t¼t b þ1

The first term represents expected lost sales, based on the margin associated with order i (mi) and a probability pðDdi Þ of the customer not placing the order, depending on the tardiness Ddi. By this, we capture that the customer may decide not to place the order after knowing its tardiness. The second term represents a fixed contractual penalty cpi for not meeting customer requirements. The last term accounts for additional (fixed) handling and shipping costs (tc), resulting from a second partial delivery. The values of Ddi, u1i ðtÞ and u2i ðtÞ are implicitly determined by the vectors x1i ¼ ðx1i ðtb þ 1Þ; . . . ; x1i ðt e ÞÞ and x2i ¼ ðx2i ðt b þ 1Þ; . . . ; x2i ðte ÞÞ; specifying the allocation of order i. As stated previously, the long-term effects, modeled by cli ðx1i ; x2i Þ, should ideally reflect the expected decrease in customer lifetime value depending on the assigned due

729

JMTM 17,6

730

date and quoted delivery quantity. If this data is not available, the long-term effects will have to be represented by an order specific penalty cost. Determining these penalty cost is, however, a difficult task for the decision maker. On the one hand, penalty costs have to reflect the relative negative effects of not fulfiling a specific customer order i0 in comparison to all other orders. On the other hand, the assigned values implicitly determine the trade-off between short term and long-term effects. If penalty costs are assigned on an order specific basis, cli ðx1i ; x2i Þ can, for example, be defined as cli ðx1i ; x2i Þ ¼ lpi · Dd i þ rpi · ð1 2 vi Þ; where the first term represents a long term penalty increasing proportionally in the tardiness of the quoted due date and the second term represents a penalty, denoted by rpi, for rejecting an order. Pre-allocation In contrast to sequential allocation, the optimization-based approach considers interdependencies between allocation decisions for individual orders within every set Oðtb Þ: With respect to future orders arriving in later periods, both sequential and optimization-based allocation mechanisms are of a myopic nature. When allocating the orders i [ Oðt b Þ to the available to promise inventory, they do not take into account that the allocation generated for Oðt b Þ impacts the ability to allocate orders arriving in Oðtb þ 1Þ; Oðtb þ 2Þ; . . . : Therefore, they cannot explicitly determine a sequence of allocations . . . ; x t0 21 ; x t0 ; x t0 þ1 ; . . . minimizing the total negative consequences b b b resulting from a stock-out situation. Pre-allocation of available to promise inventory to distinct classes of customers or individual customers can be employed to mitigate the problems associated with myopic allocation mechanisms. Commercial order fulfilment systems provide so-called “allocation planning” functions, which apply certain rules to pre-allocate available to promise quantities to sales regions and customer classes (Kilger and Schneeweiss, 2000). A typical pre-allocation scheme would, for example, employ a fixed split policy to first allocate overall available to promise quantities to sales regions and would then assign the regional quantities to individual customers or customer classes based on customer priorities. If we assume that customers in one sales region are grouped into k ¼ 1; . . . ; K pre-defined customer classes with index k reflecting an ordinal ranking based on customer priorities (with k ¼ 1 highest and k ¼ K lowest priority), we can employ the following set of rules, based on forecasted quantities f k;t for every class k, for allocating the available to promise inventory to the K customer classes:   atptk¼1 ðt b Þ ¼ min f 1;t ; atpt ðtb Þ ; ( atpkt ðt b Þ

¼ min f k;t ; atpt ðtb Þ 2

k21 X

) atpkt ðt b Þ


 ;t [ tb þ 1; te

ð17Þ


 ; ;t [ t b þ 1; te ; k ¼ 2; . . . ; K ð18Þ

k¼1

Pre-allocation of available to promise quantities can easily be combined with the allocation mechanisms described previously. Orders are then quoted from the available to promise quantities reserved for corresponding customer classes. Most commonly, allocation rules are defined in such a way that orders of higher priority can be fulfiled from quantities reserved for orders of lower priority as soon as their corresponding allotment is exhausted.

Clearly, pre-allocation is the most effective way of shifting consumption of available to promise quantities to future periods. It can prevent a system from reaching a state where all orders i with a required delivery date bd li ; dui c [ ½t b ;
t
have to be delayed or rejected. However, pre-allocation is based on highly disaggregated and uncertain data, i.e. forecasts for individual customer classes and short time buckets. Therefore, its effectiveness will largely depend on a company’s ability to determine reliable short term forecasts. A significant deviation of the ordered quantities from forecasted quantities used for pre-allocation may lead to late fulfilment or rejection of orders, although inventory quantities assigned to other customer classes would have been available. If customer priorities are used for pre-allocation, it again has to be questioned, how well they reflect the short and long term consequences of not fulfiling customer requirements.

Managing stockouts effectively

731

Case analysis Based on the data of a pharmaceutical company, we demonstrate how the different allocation mechanisms perform in a specific stock-out situation. Although, the results of the numerical analysis are obtained only for a specific instance of a stock-out situation, they provide insight into the potential of the different allocation mechanisms for managing stock-out situations. The company under consideration is a medium-sized producer of over the counter and prescription drugs, operating on a European basis. We focus on a stock-out situation for one product, cough drops, which are made to stock in large batches and sold to large and medium sized retailers, drugstore chains, and individual chemists. A stock-out situation was caused by supply shortage for one major ingredient in combination with unexpected high demand due to a wave of flu. The company was employing FCFS allocation. Figure 1 shows the planned receipts, the total quantities ordered and the available to promise quantities for the time period between October 23 and November 19. It should be noted that “ordered quantities” refer to specific due dates, and not to the dates at which orders were received. In Figure 1, the negative values represent the cumulated order quantities the company was not able to fulfil. Prior to the actual stock-out from October 31 on, the company received a number of large orders as a reaction to higher final customer demand. During the stock-out period, order quantities decreased for two reasons: due

Planned Receipts

Total QuantityOrdered

Available Quantity

Units

8000 7000 6000 5000 4000 3000 2000 1000 0 –1000 –2000

19-11

18-11

15-11

14-11

13-11

12-11

11-11

8-11

7-11

6-11

5-11

4-11

1-11

31-10

30-10

29-10

28-10

25-10

24-10

23-10

Periods

Figure 1. Characterization of the stock-out situation

JMTM 17,6

732

to large previous order quantities a number of large customers had not reached their reorder point and some of the key accounts were asked ahead of time to postpone their orders. The company was anticipating the stock-out situation one week prior to its occurrence. Therefore, it would have been possible to switch from FCFS allocation to a different allocation mechanism. As described in the previous section, the evaluation of different allocation mechanisms requires information on short- and long-term consequences of not fulfiling specific orders. The pharmaceutical company maintains long-term relationships with most large- and medium-sized customers. Typically, the companies order on a weekly or bi-weekly basis. If certain products cannot be delivered, they will most commonly be reordered the following week. Whether a stock-out of the pharmaceutical company translates into lost sales depends on the inventory held by its customers. If the stock-out situation also causes a stock-out on the customer’s side, overall order quantities will decrease. There will not be an effect on the overall sales volume if the customers carry enough inventory to fulfil their final customer demand during the pharmaceutical company’s stock-out period. The company does not posses information about customer’s inventory and lost sales caused by a stock-out situation. Contracts with some large retailers explicitly define contractual penalties if the company fails to deliver the ordered quantities according to the customer’s due date requirements. Although, the loss of a customer is not likely, the company anticipates medium- and long-term implications caused by stock-out situations. In many cases, not fulfiling customer requirements, causes a weaker position in future negotiations and leads to less favorable contractual agreements. These long-term consequences can hardly be quantified and traced to individual order promising decisions. To evaluate allocation mechanisms, a specific classification of customers was established on the basis of anticipated medium and long-term implications of order fulfilment. Account managers and sales responsibles jointly evaluated the individual customers and assigned values from 1 (very high consequences) to 5 (low consequences) to each customer. This ordinal ranking was used to establish five customer classes. For the experiments, we considered a time period of 15 weekdays (starting on October 23, six days prior to the actual stock-out situation) during which 657 orders were received. About 10.2 percent of these orders were placed by customers of class 1, 69.2 percent by customers of class 2, 5.3 percent by customers of class 3, 6.4 percent by customers of class 4 and 8.8 percent by customers of class 5. In the first set of analyses, we determined the results of rank-based allocation and optimization-based allocation with a 1 and 2-day batching interval and compared these to the results obtained from FCFS allocation. Rank-based allocation was performed on the basis of the established customer classification. To rank members of one group, the order size (both ascending and descending) was used as sub-criterion. For the optimization-based approach, we employed a modified version of the models (3)-(15). As the customers will not accept late delivery, orders that cannot be fulfiled within the required time interval have to be rejected. Optimization was based on lost margins, contractual penalties and additional transportation costs (if partial deliveries are considered) as short-term consequences of order rejection. The long-term consequences were incorporated through a fixed penalty cost depending on the customer class. As described previously, all mechanisms perform equally well as long as sufficient inventory is available to fulfil all incoming orders. Therefore, differences can only be

observed in periods, in which not all orders can be fulfiled according to customer requirements. In the specific instance considered in the experiments, the available to promise inventory was sufficient for fulfiling all orders received during the first three periods (October 25-27). In period 4 between 9 and 22, orders, depending on the employed allocation mechanism, had to be rejected. The results of the allocation for period 4 are shown in Figure 2. Because in period 4, all available to promise inventory was consumed, all orders received during the following seven periods had to be rejected. By rejecting few large orders, both rank-based (ascending) and optimization-based allocation were able to commit a larger overall number of orders. Whether this is beneficial in terms of the overall impact on the company’s profitability depends on the short- and long-term consequences associated with rejection of the individual orders. If realistic estimates for long-term consequences cannot be obtained (e.g. through customer lifetime value analysis), a company will have to compare the “rejection profiles” in order to evaluate whether the different allocation mechanisms adequately reflect their preferences. The results obtained from a comparison of the rejection profiles can also be used for determining the ratio between parameters representing short- and long-term effects. Employing rank-based and optimization-based allocation for customer orders received in period 3 and 4 (two-day batching) leads to greater discrepancies in comparison to FCFS allocation. Both approaches reduce the number of rejected orders of customer class 2 and the total number of rejections. When comparing the results, it should also be noted that only the specific sequence of order arrival during period 4 prevented FCFS allocation from rejecting orders of customer class 1. The results suggest that as short term measure, switching from FCFS allocation to a batch mode in which due dates are assigned on the basis of potential negative consequences leads to favorable results. As the effects are only observed for a small sample, we cannot generally conclude, that an optimization-based approach will yield significantly better results than a rank-based approach. As the impact of different allocation mechanisms is limited to just one period, we extended the analysis to measures shifting consumption of available to promise inventory to later periods. These include generation of partial deliveries and pre-allocation. A number of customers belonging to customer classes 2-5 accept partial deliveries. We assumed a minimum delivery quantity of 50 percent for all of the orders 1-Day Batching

(a) 25

Rejections

20

FCFS Rank (Ascending)

25 20

15

15

10

10

5

5

0

FCFS Rank (Ascending)

Rank(Descending) Optimization

0 1

2

3 4 Customer Classes

5

Overall

733

2-Day Batching

(b)

Rank(Descending) Optimization

Managing stockouts effectively

1

2

3 4 5 Customer Classes

Overall

Figure 2. Results of alternative allocation mechanisms, rejections in period 4 (October 28)

JMTM 17,6

734

placed by these customers. As concluded previously, generation of partial deliveries does not yield significant benefits in combination with FCFS. In our analysis, partial deliveries could not be generated; the remaining available inventory was not sufficient for quoting the minimum delivery quantity. Combined with a rank-based allocation, generation of partial deliveries also had insignificant impact; partial deliveries could only be generated for one order. Integrating partial deliveries into the optimization-based approach has a more significant impact on order promising. In comparison to the sequential rank-based and FCFS allocation, the optimization model specifically assigns partial deliveries if this increases overall profit. In our analysis, partial deliveries were generated for eight orders received in period 4. As suggested previously, a larger portion of inventory consumption can be shifted to future periods if partial deliveries are enforced prior to the actual stock-out situation. To analyze the impacts of this allocation strategy, two deliveries were quoted for all orders placed by customers, which are willing to accept partial deliveries. In Figure 3(a) the results of the allocation with enforced partial deliveries are compared to the results obtained from FCFS allocation. As enforced partial deliveries not only have an effect on the allocation within one single period, Figure 3(a) now shows the overall rejection rates for all 15 periods. By quoting partial deliveries for 50 orders (7.61 percent), the overall rejection rate was reduced by 53.5 percent in comparison to FCFS-based allocation. When evaluating the benefits of enforced partial deliveries it should be considered that these may also induce negative short and long term effects. In Figure 3(b), the relative number of partial deliveries and the relative reduction of rejected orders are displayed for individual customer classes. Enforced partial deliveries do not have a negative impact on orders placed by customers of classes 1 and 3. The results indicate a positive overall effect for customer classes 2 and 4. Merely for customer class 5, the relative number of partial deliveries exceeds the relative reduction of rejected orders. For this specific instance of a stock-out situation, we can conclude that enforced partial deliveries are an appropriate measure for effectively managing a stock-out situation. As pointed out before, the available to promise inventory can be pre-allocated to specific customers and customer classes. Owing to a lack of historical data and forecasting results of the pharmaceutical company, a detailed analysis of the effects of pre-allocation could not be conducted. To exemplarily illustrate the effects of pre-allocation, we employed a “naı¨ve” allocation rule: from period 1 on, all available to promise inventory was reserved for orders of customers with priority 1 and 2, all orders placed by customers of classes 3, 4 and 5 were rejected. By rejecting lower class

Figure 3. Figure 3 Results of an allocation with enforced partial fulfilment

Order Rejection 70 60 50 40 30 20 10 0

(b)

FCFS - No Partial Fulfillment Optimization - Enforced Partial Fulfillment O

Percentage

Rejected Orders (%)

(a)

1

2

3 4 Customer Classes

5

Overall

Partial deliveries vs. reduction of rejected orders 70 60 50 40 30 20 10 0

Partial Deliveries

1

2

Reduction of Rejected Orders

3 4 Customer Classes

5

Overall

orders, available to promise inventory was preserved for future orders of customers with higher priority. The results of this simple allocation led to acceptance of all orders placed by customers of classes 1 and 2. However, the amount of available inventory remaining unallocated would have been sufficient for fulfiling 52 percent of the total demand coming from orders of customer class 3. Whether or not pre-allocation is an appropriate measure for effectively managing stock-out situations can only be evaluated on the basis of results gained from a more extensive analysis. Especially allocation accuracy, the effects of forecasting errors, and the risk of rejecting orders although sufficient inventory would have been available, have to be explored. This analysis would then also have to investigate potential situations, in which inventory is allocated and orders are rejected without the anticipated stock-out situation actually materializing. Conclusions Both the general analysis of allocation mechanisms and the results obtained from the case analysis show that in a stock-out situation, FCFS is not an adequate allocation mechanism. Switching from a real-time FCFS allocation mode to a rank-based or optimization-based allocation can lead to a more beneficial allocation of orders to remaining available to promise quantities. The effects, however, are limited; the benefits can only be leveraged immediately before the available to promise quantities are consumed. If a company is able to anticipate a stock-out situation prior to its actual occurrence, it can be beneficial to shift consumption of available to promise quantities to future periods. One potential measure is the generation of partial deliveries. The general analysis and the results of the case analysis show that a major effect of partial delivery generation can only be expected if they are enforced prior to the stock-out situation. If applied shortly before available to promise inventory is exhausted, the effects are again limited. The results of the analysis indicate that this especially pertains to the case when partial deliveries are combined with FCFS and rank-based approaches. Although, (enforced) partial deliveries can effect customer satisfaction and can therefore also result in negative consequences, they can significantly mitigate the problems associated with a stock-out situation. As pointed out in the case analysis, the positive effects can more than offset negative consequences of enforced partial deliveries. In comparison to pre-allocation, partial delivery generation is less effective in terms of shifting available to promise quantities to future periods. In return, however, it does not involve the risk of unnecessarily rejecting orders. The potential of pre-allocation has not been explored extensively in this paper. A final evaluation requires more extensive analysis based on a representative set of data. References Ball, M.O., Chen, C-Y. and Zhao, Z-Y. (2004), “Available to promise”, in Simchi-Levi, D., Wu, S.D. and Shen, Z-J. (Eds), Handbook of Quantitative Supply Chain Analysis – Modeling in the E-Business Era, Kluwer, Boston, MA, pp. 447-84. Chen, C-Y., Zhao, Z-Y. and Ball, M.O. (2001), “Quantity-and-due-date-quoting available-to-promise”, Information Systems Frontiers, Vol. 3 No. 4, pp. 477-88. Chen, C-Y., Zhao, Z-Y. and Ball, M.O. (2002), “A model for batch advanced available-to-promise”, Production and Operations Management, Vol. 11, pp. 424-40.

Managing stockouts effectively

735

JMTM 17,6

736

de Ve´ricourt, F., Karaesmen, F. and Dallery, Y. (2002), “Optimal stock allocation for a capacitated supply system”, Management Science, Vol. 48 No. 11, pp. 1486-501. Fleischmann, B. and Meyr, H. (2003), “Customer orientation in advanced planning systems”, in Dyckhoff, H. et al. (Eds), Supply Chain Management and Reverse Logistics, Springer-Verlag, Berlin, pp. 297-321. Gordon, V., Proth, J-M. and Chu, C. (2002), “A survey of the state-of-the-art of common due date assignment and scheduling research”, European Journal of Operational Research, Vol. 139, pp. 1-25. Ha, A.Y. (1997), “Inventory rationing in a make-to-stock production system with several demand classes and lost sales”, Management Science, Vol. 43 No. 8, pp. 1093-103. Kilger, C. and Schneeweiss, L. (2000), “Demand fulfilment and ATP”, in Stadtler, H. and Kilger, C. (Eds), Supply Chain Management and Advanced Planning, Springer, Berlin, pp. 79-95. About the author Richard Pibernik is Professor of supply chain management at the Zaragoza Logistics Center, and a Research Affiliate at the MIT Center for Transportation and Logistics. He currently teaches supply chain management in the master’s program of the MIT-Zaragoza International Logistics Program. He received both his master’s degree and his doctorate in business administration from the Goethe-University in Frankfurt, Germany. He has active research projects on the configuration of order promising systems, supply chain process management support methodology, coordinating distributed decisions in supply chains, and configuring dynamic supply chain networks. Pibernik has presented the results of his research at various national and international conferences and has published a number of articles on specific topics in supply chain management. Richard Pibernik can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

The current issue and full text archive of this journal is available at www.emeraldinsight.com/1741-038X.htm

Optimizing supply chain management using fuzzy approach N. Gunasekaran Department of Mechanical Engineering, Kumaraguru College of Technology, Coimbatore, India

S. Rathesh

Optimizing supply chain management 737 Received July 2005 Revised January 2006 Accepted February 2006

Industrial Engineer, Fine Jewellery (I) Ltd, Mumbai, India

S. Arunachalam School of Computing and Technology, University of East London, Dagenham, UK, and

S.C.L. Koh Management School, University of Sheffield, Sheffield, UK Abstract Purpose – The purpose of this paper is to propose a fuzzy multi-criteria decision-making procedure and it is applied to find a set of optimal solution with respect to the performance of each supplier. This method with the use of Monte Carlo simulation produces overall desirability level less imprecise and more realistic than those of the conventional QFD methods for engineering design evaluation. Design/methodology/approach – A few responses obtained from customers are simulated using a triangular fuzzy QFD algorithm, Monte Carlo simulation and a multi-objective model to optimise the total user preferences. Findings – The proposed approach provides decision-making with an optimal solution less imprecise in a QFD-based collaborative product design environment. Research limitations/implications – The proposed approach depends on the few responses and the random numbers derived from simulation. The random numbers need to be used after passing them through random number testing methods. The responses obtained from the customer are considered to be genuine and original. Originality/value – The triangular fuzzy, Monte Carlo simulation and multi-objective optimisation are embedded into QFD environment to make the decisions less imprecise than that of conventional QFD and it is tested for a case study problem. It definitely helps the managers in a collaborative product design environment. Keywords Supply chain management, Fuzzy logic, Quality function deployment, Monte Carlo simulation, Decision making Paper type Research paper

Introduction In today’s competitive marketplace, companies have to develop products that satisfy customer needs to become or remain successful (Appelqvist et al., 2004). Design is a main concern in modern manufacturing companies. However, not only carrying out design will make manufacturing companies be profitable and competitive, they need to

Journal of Manufacturing Technology Management Vol. 17 No. 6, 2006 pp. 737-749 q Emerald Group Publishing Limited 1741-038X DOI 10.1108/17410380610678774

JMTM 17,6

738

improve the effectiveness of the design activity. As the inputs of engineering design evaluation are characterized by imprecise or vague requirements, they have to be represented and managed properly. For considering the above factors at the design stage, quality function deployment (QFD) tool will be employed to enhance the decision-making (Vanegas and Labib, 2001a). This paper considers the QFD planning as multi-criteria decision problem and proposes a fuzzy-based approach. The methodology proposed has overall desirability level less imprecise and more realize than those of the conventional methods. The proposed methodology incorporates a two-stage process. In the first stage, the customer’s requirement will be compared with different design requirements, with the help of fuzzified quality function deployment (FQFD). In the second stage, design requirements were prioritized by grey-based relation algorithm. The proposed methodology of grey relation algorithm on FQFD, by using Monte Carlo simulation (MCS) on fuzzy membership function helps in optimizing engineering design and also in multi-criteria decision-making. Quality function deployment The QFD is employed to determine the target values of engineering characteristics (ECs) which are translated from voice of the customer (Vanegas and Labib, 2001b). Customer requirements and their degree of importance are represented on the left side of the house of quality (HoQ). The technical design characteristics are reported on the top of HoQ. The matrix in the main body of the HoQ identifies the relationship matrix, which highlights the mutual influence between customer requirements and product engineering/design characteristics. The “roof” part of HoQ shows the correlation among technical characteristics (Franceschini and Rosseto, 2002). The right side of HoQ reports a competitive benchmarking on each customer attribute for competitor’s product. Target levels of ECs are determined by all information contained in the HoQ. Despite its apparent easiness, if information contained in the HoQ is not sufficiently “accurate” QFD can become a “misleading” tool. Its correct and effective use needs a careful design analysis and an accurate data collection (Vanegas and Labib, 2001a). After customer identification, the first step of the QFD process is the setting up of procedures for gathering information by customers. The second step concerns data management and elaboration. Typical examples of these activities are the definition of customer requirements and evaluation of their relative degree of importance. Methods for determining the importance rating of technical characteristics are dependent on the representation of the symbols contained in the relationship matrix. If symbols are converted in a 1-3-9 numerical scale, we may use the simple weighted sum method. Such procedures can become arbitrary in those situations in which the customer is not able to give a significant evaluation of his requirements and his preference system is not explicitly known. The extreme consequence of the use of inadequate conversion can lead to a setting up of a design of a product for an “ideal” customer, which is different from the real one. The soft issues are that we do not know the “distance” between the tow designs. With specific reference to QFD, the introduction of weights to assign a relative degree of importance to customer requirements can lead to a prioritization order of technical characteristics, which does not reflect customers’ real intentions (Franceschini and Rosseto, 2002).

The ranking of technical design requirements The QFD approach provides two steps for the ranking of technical design characteristics. The first one concerns the artificial conversion of the relationship between customer requirements and design characteristics into numerical equivalent values (Franceschini and Rosseto, 2002). A special score is obtained by substituting nine points for a strong relation (symbol ;), three points for a moderate relationship (symbol †), and one point for a week relationship (symbol D). Numerical values so obtained represent the new coefficients of the relationship matrix R as shown in Figure 1. The second step provides the determination of relative weights (RWs) w0j of technical design characteristics: w0j ¼

K X

di · r ij ;

j ¼ 1; 2; . . . ; n:

Optimizing supply chain management 739

ð1Þ

i¼1

where: di ¼ degree of importance of the customer requirements i-th, i ¼ 1; 2; . . . ; m; rij ¼ numerical relationship between the customer requirements i-th and technical design characteristics j-th; i ¼ 1; 2; . . . ; m; j ¼ 1; 2; . . . ; n; w0j ¼ importance rating for the technical design characteristics j-th; j ¼ 1; 2; . . . ; n; m ¼ number of customer requirements; n ¼ number of technical design characteristics. Relative importance weights are obtained as follows: wj ¼

w0j ; n X 0 wj

j ¼ 1; 2; . . . ; n:

ð2Þ

Brightness Linearity Focus White Balance Convergence Absolute weight Relative weight (%)

Total Absolute weight

Electron Tube

Electron Gun Sealing

Fluorescent Paint

Monitor Users Requirement

Black Matrix

CRT Design Process

Mask Operation welding

Degree of importance

j¼1

9 3 9 1 3 111 87 27 135 144 504 22 17.3 5.36 26.8 28.6

Figure 1. The HoQ for CRT example using traditional QFD approach

JMTM 17,6

740

Weights so determined represent the importance that the customer indirectly describes to each design characteristics. They can be interpreted as the degree of “attention” that a designer must reserve to each single technical characteristic during the product development process. The determination of weights by means of equation (1) needs the knowledge of the degree of importance of each customer requirements (dj), and the conversion of symbols contained in the relationship matrix into “equivalent” numerical scores (rij). The shaded region shown in Figure 1 is the prioritized HoQ for technical characteristics and customer needs for the example listed above. Fuzzy QFD (FQFD) An important feature in QFD is that human input is used to determine both the importance of each performance aspects and the relationship scores. Traditional QFD requires humans to translate their perceptions into numerical scales (Vanegas and Labib, 2001b). For example, the human is asked to denote if a relationship is low, high or very high and their answer is translated to a scale like 1-3-5, 1-3-9 or 1-5-9. Atleast two problems with this approach potentially degrade the value of a traditional QFD study. The first problem is that not everyone has the same perception of a particular linguistic description. For example, response of “high” from several different people does not typically mean the same thing, yet they are assigned the same score in traditional QFD. The fact is that perception of moderate to one person is equivalent to strong in the mind of another. Failing to capture this ambiguity can create a bias in the QFD result. The second problem is that the choice of scales can dramatically influence the outcome. The fuzzy QFD approach proposed in this paper is to determine the optimal levels of requirements to satisfy each customer. Fundamentally, fuzzy arithmetic (Jain et al., 2004; Lowen, 1996) is applied to the linguistic expression to describe the importance of each performance aspect, in order to construct a quantitative measure of these responses that combines the inherent ambiguity of linguistic statements. Fuzzy QFD begins with human input from a number of experts using one of a finite number of linguistic variables (LV), LVk, k ¼ 1; 2; . . . ; k: For example, suppose the experts are used to characterize the relationship between a particular technical characteristics and customer needs as strong, moderate, and weak. These are assigned the notation LV1 ¼ strong (S), LV2 ¼ moderate (M), LV3 ¼ weak (W). Each of these descriptors is treated as a fuzzy set, bounded to a predetermined interval, and characterized for this example in the interval [1,9]. Figure 2 shows one set of assignments that can be made. µ Ã (x)

Weak

Moderate

Strong

1.0

0.5

Figure 2. Membership functions for LV

0 1

3

9

x

~ relates the possible quantitative values; the A membership function of a fuzzy set A linguistic response may take a probability for each value selected. For example, if ~ is x; a; b; c [ R; a , b , c, and R ¼ (2 1 1), the membership function, mAðxÞ; defined as (Lowen, 1996): 8 ðx 2 aÞ=ðb 2 aÞ; a # x # b; > > < ~ ð3Þ ¼ ðc 2 xÞ=ðc 2 aÞ; b # x # c; mAðxÞ > > : 0; otherwise:

Optimizing supply chain management 741

Fuzzy QFD using MCS and appropriate membership function is used to quantify each response and then the responses from all experts are averaged to provide the final value of the measure (Abebe et al., 2000). To illustrate, consider the membership function shown in Figure 2 that are mathematically defined in Table I. Now suppose M experts provide their opinion on the relationship between customer needs i and technical characteristics j using one of the k LV provided. The relationship score is calculated by: M X

S ij ¼


 MCS triang LVm k

m¼1

ð4Þ

; i and j;

M

where MCS½triangðLVm k Þ
is the MCS value for the appropriate membership function of the linguistic variable LVk that was chosen by expert m. The same process is repeated for each customer needs – technical characteristics to generate the entire set of relationship scores. The importance of each performance aspect is computed similarly. Grey-based fuzzified QFD The grey relationship grade represents the degree of relation between the reference sequence say xo, and the comparison sequence say xi. The higher degree of relation means the comparison sequence is more similar to the reference sequence than for other comparison sequences. This measurement can be easily applied to measure the similarity between data. By using the equation (4), relationship score for each technical characteristics and customer needs can be calculated. After calculating relationship score (Sij), the relationship between each relationship score value for each customer response can be measured by using grey relation formula. This will give out less imprecise value compared to other methods, the value obtained is used further to calculate the degree of importance: Gðxo ; xi Þ ¼

Dmin þ zDmax ; Doi þ zDmax

Fuzzy set

Membership function

Domain

Strong Moderate

U(x) ¼ (x 2 9)/(9 2 3) U(x) ¼ (x 2 1)/(3 2 1) U(x) ¼ (9 2 x)/(9 2 3) U(x) ¼ (3 2 x)/(3 2 1)

3#x#9 1#x#3 3#x#9 1#x#3

Weak

ð5Þ

Triangular (min, mode, max) 3, 9, 9 1, 3, 9 1, 1, 3

Table I. Fuzzy set and membership functions

JMTM 17,6

742

where z [ [0,1] is the distinguished coefficient and Doi, Dmax and Dmin are, respectively, described as: Doi ¼ jxo 2 xi j

ð6Þ

Dmax ¼ maxjxo 2 xi j

ð7Þ

Dmin ¼ minjxo 2 xi j

ð8Þ

Method of total preferences The method of total preference performs its function by using the RW, RW’s with additional human expert’s opinion to develop a single measure that reflects the ability of each alternative to satisfy the performance aspect, this is defined as the total user preference (TUP) for alternative n, TUPn, and is computed as: TUPn ¼

J X

RWj WAnj

; n;

ð9Þ

j¼1

where RWj is the relative technical importance rating of technical requirements j (WAnj) the degree to which alternative n can deliver technical requirements j. Before computing the TUPs the degree to which alternative n can deliver requirements j, WAnj must be determined. The proposed methodology determines the WAs for input provided by the decision maker using fuzzy QFD: M X

WAnj ¼ m¼1


 MCS LVm k M

; n and j:

ð10Þ

Finally, the normalized TUP for alternative n (NTUPn), is calculated by: NTUPn ¼

TUPn N X TUPn

; every n;

ð11Þ

n¼1

A summary of these calculations for a prototype problem is illustrated in Table II. Multi-objective model The final stage of the proposed methodology is the multi-objective model. Clearly, no single model is capable of being applicable to all situations rather a model must be built that represents the situation. In the following, two objectives are considered that we assume conflict, namely, maximizing customer value as measured by user satisfaction and minimizing cost. The objective functions shall be formulated as follows: P (1) Minimize TC ¼ ni¼1 TCi X i P (2) Maximize TUS ¼ ni¼1 NTUPi X i

Req. 1 Req. 2 .. . Req. j TUPn NTUPn

Alternative 1

Alternative 2

Alternative 3

...

Alternative N

RW1* WA11 RW .. 2* WA12 . RWj* WA1j n X RWj *WA1j i¼1 n X TUP1 = TUPn

RW1* WA21 RW .. 2* WA22 . RWj* WA2j n X RWj *WA2j i¼1 n X TUP2 = TUPn

RW1* WA31 ..RW2* WA32 . RWj* WA3j n X RWj *WA3j i¼1 n X TUP3 = TUPn

... .. . . .. ...

RW1* WAN1 RW .. 2* WAN2 . RWj* WANj n X RWj *WAij i¼1 n X TUPN = TUPn

i¼1

i¼1

i¼1

... ...

i¼1

where TUS is the total user satisfaction, TC the total cost, TUPn the TUPs for vendor n, TCn the total cost of vendor n and: ( 1 if vendor n is selected; Xn ¼ 0 otherwise: To resolve this problem, a strategy that parallels preemptive goal programming is employed. A sub-problem is formed and solved with the highest priority objective as the sole objective and a constraint is added to maintain the secondary objective at a predetermined threshold level specified by the user. The first sub-problem solved as: max TUS ¼

n X

NTUPi X i

i¼1

such that

n X

X [ S; X n [ ð0; 1Þ for every n;

TCi X i # TCmaximum ;

i¼1

where TCmaximum is the maximum threshold value of TC, S the feasible region. A solution to this problem yields the maximum value of TUS for the threshold value of TC. A second sub-problem is now formulated and solved in which TUSminimum is used as the threshold value of TUS and TC is minimized. The most obvious initial value of TUSminimum is the optimal value formed by solving the first sub-problem: Min TC ¼

n X

TCi X i

i¼1

such that

n X

NTUPi X i $ TUSminimum ;

X [ S; X n [ ð0; 1Þ for every n;

i¼1

The resulting solution represents the best choice for the decision-maker. The model is actually used as a quantitative tool to support this complex decision-making process rather than providing the solution. To facilitate this, the second sub-problem is repeatedly solved as the threshold value for the primary objective is incrementally relaxed. In the above example alternative solutions would be generated by repeatedly

Optimizing supply chain management 743 Table II. Total user preferences

JMTM 17,6

744

solving second sub-problem as the value of TUSminimum is decreased. The decision maker is then presented a series of compromise solutions from which he or she can proceed to impose other decision criteria. An illustrative example Let us consider the simple case of a design of cathode ray tube (CRT) with supply chain network. The authors have made an attempt to determine the technical design characteristics prioritization form the customer point of view. Figure 3 shows the supply chain network with reliability as one major factor. Figures 4 and 5 show the HoQ for the CRT product. It represents, the degree of importance is based on the number of responses from customers as shown in Table III. Importance of the performance aspect Using the information from Table III, the importance of each performance aspect is computed as: Customers Reliability Requirement & Claim

Electron gun Glass vendor

Figure 3. Reliability consideration in SCM

CRT Mfg.

TV Set Mfg

End User

Internal Reliability Assurance Testing

Figure 4. The HoQ for CRT example using fuzzy QFD approach (stage-I)

6.66 3.99 6.34 1.66 4.07

5.67 2 4.68 4.33 4.33 1.93 75.4 42.5 21.7 12.2

2 3.98

Total Absolute weight

Electron Tube

Electron Gun Sealing

Fluorescent Paint

Black Matrix

Mask Operation welding

Monitor Users Requirement Brightness Linearity Focus White Balance Convergence Absolute weight Relative weight (%)

Degree of importance

CRT Design Process

3.99 5 3.01 6.99 5

4.35 6.67 3.01 46.9 79.5 103 347.4 13.5 22.9 29.7

n h  io dbrightness ¼ 3 MCS triang LVm ; strong 
 ¼ 7:007; 6:986; 7:003; 4:314 MCS triang LVm moderatek The value of 7.007, 6.986, and 7.003 are obtained from MCS using triangular distribution for “Strong” shown in Figure 2 while 4.314 is the value from the distribution for “Moderate”.

Optimizing supply chain management 745

Grey-based algorithm for finding degree of importance Using equation (5) the degree of importance for each performance aspect are calculated as: Gð7:007; xi Þ ¼

0:004 þ 0:3ð2:693Þ ¼ 0:9794; 0:021 þ 0:3ð2:693Þ

Similarly grey relation is found for all pairs of importance of the performance aspect, and the pair’s having highest relation are taken for averaging as show below:

4.33 4.31 6.33 2 6.98

Total Absolute weight

STEM - C

PIN - C

MIX

SHORT

COEK

CRT Design Process Mask Operation welding Black Matrix Fluorescent Paint Electron Gun Sealing Electron Tube Absolute weight Relative weight (%)

Degree of importance

Aging Test

6.67 2 3

1.66 4.35 4.33 7 5.66 3 4.99 3.99 7.01 7.01 4.01 5.67 3 105 7.98 96.5 65.8 102 377.2 27.9 2.12 25.6 17.4 26.9

Monitor users Req.

Number of responses

Brightness Linearity Focus White balance Convergence

3 3 3 4 2

strong, 1 moderate moderate, 1 weak strong, 1 weak weak moderate, 1 weak, 1 strong

Figure 5. The HoQ for CRT design – aging test using fuzzy QFD approach (stage-II)

Table III. Number of responses

JMTM 17,6

Gbrightness ¼

½ð7:007 þ 7:003Þ þ ð6:986 þ 7:003Þ þ ð7:003 þ 7:007Þ þ ð4:314 þ 6:986Þ
8

¼ 6:66:

746

Relationship scores Based on the customers opinion for the relationship between each customer needs and technical characteristics. The relationship scores are calculated, suppose two members think that relationship between Brightness to Mask operation and welding as “Strong” and remaining two members thinks as “Moderate”. Then Sbrightness – Mask operation and welding is calculates as: S brightness 2 X

Mask operation and welding

2 h  i 
 X þ MCS triang LVm MCS triang LVm moderate strong

¼ m¼1

m¼1

4

¼ 5:67: The final transformation matrix and degree of importance are shown in Figures 4 and 5. The method of total preferences Before computing the TUPs from equation (9), the WAs must be determined using FQFD for each requirement – vendor pair. For, e.g. WAMask operation and welding2V 1 ¼ 4:307: Hence, TUPMask operation and welding2V 1 ¼ RWMask operation and welding £ WAMask operation and welding2V 1 ¼ 0:217 £ 4:307 ¼ 0:935: The values of TUP for each vendor are in Table IV. Vendors Criterion

Table IV. NTUPs and TUPs value

Mask operation and welding Black matrix Fluorescent paint Electron gun sealing Electron tube COCK SHORT MIK PIN – C STEM – C Total user preferences Normalized TUP

V1

V2

V3

V4

V5

V6

0.935 0.528 0.225 1.604 2.091 1.216 0.035 1.807 0.758 1.895 11.09 0.184

0.935 0.854 0.225 1.613 1.285 1.207 0.035 1.792 0.291 1.167 9.404 0.156

1.532 0.525 0.584 1.616 2.081 0.461 0.092 1.106 0.755 1.883 10.64 0.177

0.927 0.855 0.225 1.605 0.495 1.206 0.035 1.113 1.217 1.889 9.567 0.159

0.942 0.531 0.225 1.606 1.282 1.215 0.035 1.796 0.755 1.883 10.27 0.171

0.938 0.854 0.225 1.606 2.082 0.466 0.092 1.794 0.752 0.448 9.257 0.154

Multi-objective modeling For this example, it is assumed that the costs of the product by alternative vendors, as per the details in Table V. The team decides that NTUPs is the highest priority objective and TC is the secondary objective; however, TC must not exceed Rs. 5,000. Since, the NTUP is the primary objective, the first sub-problem is: Max TUS ¼

6 X

Optimizing supply chain management 747

NTUPi X i

i¼1

Such that

6 X

TCi X i # 5; 000;

X [ S; X n [ ð0; 1Þ; i ¼ 1; . . . ; 6

i¼1

The optimal solution to this problems is TUS * ¼ 0.184 for X * ¼ vendor 1. The TC for vendor 1 is Rs. 4,500. Solving the second sub-problem, minimize TC while keeping TUs . 0.184 yields vendor 1 as optimal, so the minimum value of TUS is now relaxed. The first level considered is TUSminimum ¼ 0.154, so the problem becomes: Min TC ¼

6 X

TCi X i

i¼1

Such that

6 X

NTUPi X i $ 0:154;

X [ S; X n [ ð0; 1Þ; i ¼ 1; . . . ; 6:

i¼1

The optimal solution for this problem is TC * ¼ Rs. 2,150, X * ¼ vendor 3 and the associated TUS ¼ 0.177. By further relaxing the TUSminimum, alternative solutions are generated, some of which are illustrated in Table VI. The NTUP and TC decide the vendor to be selected. Even though, the vendor 6 quotes less cost, the NTUP is significantly less. And hence, the proposed methodology provides support in decision-making for selecting his vendor. Conclusion To systemize the engineering design process, reduce cost and leading time, many methods are emerging now a day, which are different from the conventional one. Vendor 1 Vendor 3 Vendor 5

Rs. 4,500.00 Rs. 2,150.00 Rs. 3,750.00

Vendor 2 Vendor 4 Vendor 6

NTUP

TC

0.184 0.177 0.154

Rs. 4,500.00 Rs. 2,150.00 Rs. 2,650.00

Rs. 6,700.00 Rs. 5,350.00 Rs. 2,650.00

Table V. Cost of each vendor’s software

Selected vendor Vendor 1 Vendor 3 Vendor 6

Table VI. Alternative solution to illustrative example

JMTM 17,6

748

Fuzziification in the evaluation of design stage, is becoming a necessary element in the engineering design process, as most of the information managed such as importance of customer requirement and prioritized technical characteristics which make it clear that fuzzy numbers are more appropriate than subjective crisp numbers. The methodology described in this paper is a quantitative tool designed to assist decision makers, when they are faced with a problem that requires selecting from a finite number of alternatives, contain multiple objectives that conflict. And also it helps to know the important functionalities for “cathode ray tube” which involves data that is both tangible like cost and intangible like LV. This methodology enhances the decision-making in a collaborative product design environment. The use of fuzzy QFD also brings a sense of total user focus to the process. Further with the output including a set of solution along with the degree of compromise associated with each relative to the different objective, the team is encouraged to address priority in quantitative fashion. Finally, the methodology will assist in pinpointing those areas of concern where the team involvement and the use of specified tools are most needed. Implications and future research The outcome of this approach depends on the few responses received from customers and simulation based on random numbers. Hence, it should be applied where there is a need. The random numbers could be tested and used to avoid any unforeseen situation with regard to the random numbers. Few more real life case studies could be conducted to ensure the readiness of this approach for any other problem. References Abebe, A.J., Guinot, V. and Solomatine, D.P. (2000), “Fuzzy alpha-cut vs Monte Carlo techniques in assessing uncertainty in model parameters”, Proc. of 4-th International Conference on Hydroinfomatics, Iowa City. Appelqvist, P., Lehtonen, J.M. and Kokkonen, J. (2004), “Modelling in product and supply chain design: literature survey and case study”, Journal of Manufacturing Technology Management, Vol. 15 No. 7, pp. 675-86. Franceschini, F. and Rosseto, S. (2002), “QFD: an interactive algorithm for the prioritization of product’s technical design characteristics”, Integrated Manufacturing Systems, Vol. 13, pp. 69-75. Jain, V., Tiwari, M.K. and Chan, F.T.S. (2004), “Evaluation of the supplier performance using an evolutionary fuzzy-based approach”, Journal of Manufacturing Technology Management, Vol. 15 No. 8, pp. 735-44. Lowen, R. (1996), Fuzzy Set Theory, Kluwer Academic Publishers, London. Vanegas, L.V. and Labib, A.W. (2001a), “Application of new fuzzy-weighted average (NFWA) method to engineering design evaluation”, International Journal of Production Research, Vol. 39, pp. 1147-62. Vanegas, L.V. and Labib, A.W. (2001b), “A fuzzy quality function deployment (FQFD) model for deriving optimum targets”, International Journal of Production Research, Vol. 39, pp. 99-120. Further reading Bojadziev, G. and Bojadziev, B. (1997), Fuzzy Logic for Business, Finance and Management, World Scientific Publication, Singapore.

Erol, I. and Ferrel, W.G. Jr (2003), “A methodology for selection problems with multiple conflicting objectives and both qualitative and quantitative criteria”, International journal of Production Economics, Vol. 86, pp. 187-99. Khoo, L.P. and Ho, N.C. (1996), “Framework of a fuzzy quality function deployment system”, International Journal of Production Research, Vol. 34 No. 2, pp. 299-311. Kumar, R. and Midha, P.S. (2001), “A QFD based methodology for evaluating a company PDM requirements for collaborative product development”, Industrial Management & Data Systems, Vol. 10 No. 3, pp. 126-31. Ohdar, R. and Ray, P.K. (2004), “Performance measurement and evaluation of suppliers in supply chain: an evolutionary fuzzy-based approach”, Journal of Manufacturing Technology Management, Vol. 15 No. 8, pp. 723-34. Power, D. (2005), “Implementation and use of B2B-enabling technologies: five manufacturing cases”, Journal of Manufacturing Technology Management, Vol. 16 No. 5, pp. 554-72. Ross, T.J. (1995), Fuzzy Logic with Engineering Applications, McGraw-Hill International Edition, Singapore. Wang, J. (1999), “Fuzzy out ranking approach to prioritize design requirements in quality function deployment”, International Journal of Production Research, Vol. 37 No. 4, pp. 899-916. Weber, C., Werner, H. and Deubel, T. (2003), “A different view on product data management/product life-cycle management and its future potentials”, Journal of Engineering Design, Vol. 14 No. 4, pp. 447-64. Corresponding author N. Gunasekaran can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

Optimizing supply chain management 749

The current issue and full text archive of this journal is available at www.emeraldinsight.com/1741-038X.htm

JMTM 17,6

750

A knowledge-based service automation system for service logistics C.F. Cheung, Y.L. Chan, S.K. Kwok, W.B. Lee and W.M. Wang Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, People’s Republic of China Abstract Purpose – Effective service logistics can lower the cost and increase service value by improving customer satisfaction and loyalty. However, the conventional ways of the service logistics are information driven instead of knowledge-driven which are insufficient to meet the current needs. The purpose of this paper is to present a knowledge-based service automation system (KBSAS) to enhance the competitiveness for manufacturing enterprises in service logistics. Design/methodology/approach – The KBSAS incorporates various artificial intelligence technologies such as case-based reasoning which is used for achieving four perspectives of knowledge acquisition, service logistics, service automation and performance measurement, respectively. Findings – A prototype customer service portal has been built based on the KBSAS and implemented successfully in a semi-conductor equipment manufacturing company. It is verified that the KBSAS provides high quality customer services with fast and efficient customer responses. It also allows the company to capture the valuable experience and tacit knowledge of the staff in performing customer and field services. Practical implications – The KBSAS yields a number of advantages over conventional service logistics which include streamlining the service logistics process; performance measurement; reduction of paper work; the provision of 24 hours worldwide automatic customer service supported by the verified knowledge base established in the date time operations as well as the driving for continuous improvement of customer service quality. Originality/value – The paper presents the development and successful implementation of a KBSAS which allows for the capture of the valuable experience and tacit knowledge of the staff in performing customer and field services. Keywords Artificial intelligence, Customer service management, Customer relations, Knowledge management Paper type Research paper

Journal of Manufacturing Technology Management Vol. 17 No. 6, 2006 pp. 750-771 q Emerald Group Publishing Limited 1741-038X DOI 10.1108/17410380610678783

1. Introduction Manufacturing organizations are now facing shortening product life cycle, growing emphasis on process automation, global competition, and an increasingly mobile workforce. There is a need for robust methods for service automation and service logistics so as to minimise the downtime in the plants and to retain the valuable operation and management knowledge within the enterprise (Lee, 2003; Zhang et al., 2003; Jiang et al., 2002). Apart from this, customer satisfaction is one of the important The authors would like to express their sincere thanks to the Research Committee of The Hong Kong Polytechnic University for financial support of the research work (Project Code: A-PG01).

factors, which affect the success of an enterprise (Feinberg and Kadam, 2002). Customer relationship management (CRM) is an integration of technologies and business processes used to satisfy the needs of a customer (Bose, 2002; Anderson and Kerr, 2002). In CRM, customer service management (CSM) is a vital part, which enables customers to individually monitor and control their service (Langer et al., 1999). Although the concept of CSM has been used in different industries such as banks, insurance and telecommunications, the advantages of CSM systems are generally overlooked in manufacturing industries. This is particularly truth for high-tech companies such as the semiconductor equipment manufacturers. To achieve effective and efficient CSM in manufacturing organizations, service logistics is indispensable which allows the delivery of the right quantity, fit and mix of spare parts and maintenance service for the customers, end-users, distributors and site engineers when they need them. Superior levels of service value and performance are driven by the service logistics strategies. This is accomplished by managing the product life cycle and new product introduction to service; forecasting the needs based on historical usage of the spare parts; acquiring parts and services, spare-part logistics (SPL), and managing the supply chain. However, the conventional ways of the service logistics are information driven instead of knowledge-driven which are insufficient to meet the current needs. For example, spare parts are located in different places and parts are transferred between customer sites and warehouse. However, the capabilities and knowledge needed to provide the maintenance service, track the spare parts, field service workforce management, supply chains, contracts, and repair are available in logistics and enterprise resources planning (ERP) systems. Since, the knowledge in performing customer and field maintenance services are difficult to be acquired, shared and diffused among the staff, training up of well-experienced staff are not only time consuming but also expensive. The enterprise would be suffered from the lost of the valuable know-how and tacit knowledge of the staff when he or she left the enterprise. As a result, this paper presents a knowledge-based service automation system (KBSAS) for service logistics. The capabilities of the system were evaluated through the trial-run implementation at a selected reference site. 2. Conventional vs knowledge-based service automation approaches 2.1 Conventional approach As shown in Figure 1(a), the conventional approach of service logistics starts with receiving customer requests by phone, fax or email. Based on the past records, the customer service staff needs to verify the identity of the customer and record the requests. Based on the experience, knowledge and availability of the field service personnel and spare parts, an appropriate field service personnel is assigned and spare parts are allocated for providing on-site service to the customer. After providing the on-site services, the failed parts are returned through a reverse-logistics process for depot repair, refurbishment, and upgrade. Typically, most of spares inventory is recycled, as failed parts are brought back to depot repair for refurbishment and reuse. Service reports are then written by the field service personnel. The customer service staff contacts the customers for the verification of the reports. The performance of the field service personnel is evaluated by the customers using the feedback questionnaire.

A knowledgebased system

751

JMTM 17,6

(a)

(b)

Customer Request

Customer Request

Realizes the Problems

752

Retrieve the past service records

Update the Service Records

KBSAS

Allocation of Field Service Personnel Spare Part logistics

Conduct on-site service at the customer site

Figure 1. A comparison between (a) the conventional approach; and (b) the knowledge-based automation approach for service logistics

Reporting

Conduct on-site service at the customer site

Reporting

Evaluate the

Evaluate the

Customer Service

Customer Service

Quality

Quality

The quality of service provided by the field service personnel is assessed by the service manager based on the customer feedback. In the conventional approach, it involves complicated knowledge flow and a lot of paper work is needed to record the orders, report the cases and measure the service quality. As a result, useful information may be missed due to human error. After receiving the orders from the customers, a lot of time is needed for the staff to search for the relevant information and knowledge to address the customer’s requests. Since, the experience and skill are tacit knowledge, it is difficult to share the knowledge among the staff. It is interesting to note that the quality of customer services heavily rely on the know-how, experience and quality of the staff. The enterprise may suffer from the lost of the valuable know-how and tacit knowledge of the staff when he or she left the enterprise. Moreover, the spare parts demand is highly variable and depends upon many factors such as failure rates, product lifecycle, etc. There is a need to take into account the returned parts, warranties, items sent to the vendor for repair, items awaiting repairs and failed items to be received from various customer’s sites and field service personnel. Moreover, the customer service can only be provided during the office hours. When the problems are raised after the office hours, the customers cannot be served immediately and the customer loyalty may be affected.

2.2 Knowledge-based service automation approach For the knowledge-based service automation approach as shown in Figure 1(b), the processes of service logistics as well as the management of the maintenance service are accomplished by a KBSAS. The KBSAS is built for achieving four perspectives, which include knowledge acquisition perspective, service logistics perspective, service automation perspective and performance measurement perspective, respectively. The knowledge acquisition perspective refers to the assimilation of the knowledge and experience of the field service personnel and customer service staff in the response of the customer’s requests which is accomplished with the use of a knowledge-based system (KBS) based on a verified set of past successful cases. The knowledge captured during daily operations can be used to automate the service off the office hours, i.e. the service automation perspective. Field service management (FSM) and SPL are undertaking for the service logistics perspective. The performance of the service staff can also be continuously measured and evaluated so as to achieve the performance measurement perspective.

A knowledgebased system

753

3. Architecture of the knowledge-based service automation system An architecture of the KBSAS is shown in Figure 2 which is established for achieving four perspectives as described in Section 2. The KBSAS is basically composed of five components, which are knowledge repositories (KR), a KBS, a service logistics module (SLM), a performance measurement module (PMM) and service automation module (SAM), respectively. The KR is composed of case library, databases and dynamic data for supporting the other modules. The assimilation and the diffusion of the knowledge of the diverse service logistics activities are accomplished by the KBS. The SLM serves the Service Logistics Perspective

Knowledge Acquisition Perspective

Performance Measurement Perspective

Field service personnel

Customer Request

Management staff

Field service management and spare part logistics

Customer Service Staff Intranet Interface

Management Staff Intranet Interface

Service logistics module (SLM)

Spare part inventory, field service schedule, customer request, etc.

Knowledge-based System(KBS)

Knowledge Repositories (Verified Cases)

Performance Measurement Module (PMM)

Service Automation Perspective

Enterprise perspectives

Internet Customer Request

Users perspectives

Interface for Extranet/Internet Customers

Service Automation Module (SAM)

Verification of cases

Portal interface

Back-end integrated applications Figure 2. Architecture of the knowledge-based service automation system (KBSAS)

JMTM 17,6

purposes of automating spare parts logistics and optimising the schedule and resources for the field service staff to provide onsite service to the customers based on the knowledge assimilated by the KBS. For the SAM, the knowledge captured during daily operations is used to support the customer service process off the office hours via the internet. The service quality of the staff is measured and monitored continuously by the PMM.

754

3.1 Knowledge repositories The KR consists of case libraries and the databases for storing the information related to the customer, the field service staff, the customer privileges, business information, reasoning list, parameters for measuring the performance of the system. These data are shared by all other modules. The case library is made up of a number of previous cases, which are stored in an object-oriented structure in the case database. Any cases in the case library are made up of three parts, which are the case number, case indexes and the solution sets. Structure of each case can be represented as follows: 3.1.1 Case (case number, case indexes, solution sets). The case number is assigned by the system sequentially which provides the unique designation of individual case. The case indexes are the identities of the cases that are entered into the case library so as to ensure that they can be accurately retrieved. As different customers may have different privileges for receiving services, the recommendations from the KR may vary for different customers. For example, a high-valued customer deserves for better service. As a result, the customer’s privileges are associated with the stored cases during the retrieval process. This provides an important means for the CSM. 3.2 Knowledge-based system A schematic diagram of the knowledge-based system (KBS) is shown in Figure 3. The KBS is built based on case-based reasoning (CBR) which is a recent approach to problem solving and learning that has got a lot of research interest over the last decades (Aamodt and Plaza, 1994; Watson and Marir, 1994; Finnie and Sun, 2003; Simoudis, 1992). CBR provides a self-adaptive capability for the acquisition, reuse, share, and diffuse the implicit experience and tacit knowledge of the service staff in inferencing and making recommendations for the customer request. Basically, a customer service case is composed of explicit and tacit elements, respectively. The explicit elements refer to those optional items, which have a range of well-defined choices for the staff to select (e.g. check boxes or combo box, etc.). For the explicit elements, the decision tree analysis is used to extract the relevant cases. The decision tree is composed of a series of rule-based filters, which can filter out those irrelevant cases, which do not match with the criteria of the decision rules in a decision tree. For tacit elements, similarity analysis is used to select the most similar cases based on a nearest neighbour algorithm (Cheung et al., 2003). The similarity of a case is determined as follows: m X

Similarity ¼

j¼1

wj simðvoj ; vrj Þ Pm

j¼1 wj

ð1Þ

where m is the number of inputs, wj is the weighting of the jth input, voj and vrj are values of the jth inputs and that for the retrieved cases, simðvoj ; vrj Þ is the similarity function for the jth inputs as follows:

A knowledgebased system

Case-Based Reasoning Model Transaction Database

Customer Request

755 Past Case 1

Past Case 2

New Case

Decision Tree Analysis

Weighting System

Reasoning Weighting Database

Weighting of Criteria

Past Case 3

Score

Value of Criteria

Similarity Analysis

Retaining Case

Inference Engine

Confirming Case

Recommended Solution

Adapting Case

Figure 3. A schematic diagram of the KBS

Number of Criteria

simðvoj ; vrj Þ




vo 2 vr

j j
¼12
o

jvj j þ jvrj j


ð2Þ

The retrieved cases are ranked in descending order according to the similarity. The most similar cases are then selected and suggestions are deduced by an inference engine. It is used to assess the current situation and deduce new knowledge and information to be used in reasoning process. The recommendation of solutions is accomplished by some sort of inferencing techniques such as forward-chaining, backward-chaining, fuzzy set theory, etc. In the present study, the forward chaining inference engine is used for the deduction of the solutions. By applying different kinds of rules in the inference engine, recommendations are then made by the CBR algorithm. A series of index for the recommendations are determined and the corresponding descriptions of the recommendations are retrieved from the reasoning database based on the given indexes. The recommended solution is then selected and adapted by staff to resolve the customer’s problems. The recommended case may need to be revised to fit the customer needs. Then the revised case is retained as new case in the KR. The description together with the solution of the confirmed case is then stored in the KR for future reuse.

JMTM 17,6

756

3.3 Service logistics module The SLM is composed of SPL component and FSM component, respectively. The SPL component serves the purposes of handling spare parts logistics which include the storage and distribution of spare parts to the aftermarket, spare-part inventory management as well as deriving useful knowledge from ERP system. In spare parts logistics that support field service operations, parts flow down to the field and then return through a reverse-logistics process for depot repair, refurbishment, and upgrade. This reverse-logistics process changes the demand forecast and the optimum supply sources and rate. The derived knowledge is used to accomplish the spare part replenishment decision, etc. As shown in Figure 4, the SPL component is built based on a dynamic forecasting model (Cheung et al., 2005) and a CBR model. Effective inventory management of spare parts demands for an accurate and dynamic forecasting model. It is vital to ensure that the required spare parts are available for field service operation while the inventory holding of the spare parts can be kept to the minimal. However, conventional ways of forecasting based on moving average, regression analysis or exponential smoothing are neither accurate nor robust enough to response and adapt dynamically to the rapid market changes (Kress and Snyder, 1994). Furthermore, the underlying statistics of the market information change from time to time, which demands for highly adaptive forecasting model, which is robust enough to response and adapt well to the fast changes in the data characteristics. Spare Part Request

ERP inventory record

Reverse-logistics process

Check for Availability

Knowledge Repository (Business rules and verify set of conditions)

Inference Engine

Advisory Recommendation

Adapt/Revise Solutions

Figure 4. Schematic diagram of SPL component

Vendor inventory

Decision

CBR Cycle

Retained Case

Verify and Retain Case

In the present study, the forecast of the spare-part demand is accomplished by a dynamic forecasting model. The dynamic forecasting model starts with the forecast of the data by an autoregressive (AR) time-series model (Widrow and Hoff, 1960). For a nth-order AR time-series, the predicted data value y_ðkÞ is expressed as a linear combination of the n previous values i.e.: X _ ai ðkÞyðk 2 i Þ y ðkÞ ¼ N ðkÞ þ

757

n

ð3Þ

i¼1

where N(k) is the white noise, y(k) is sampled data at time index k and ai(k) are the time-series coefficients. A nth-order time-series coefficient vector is defined as: aðkÞ ¼ ½a1 ðkÞ; a2 ðkÞ; . . . ; an ðkÞ
T

ð4Þ

A vector of current and n 2 1 past data is defined as: yðkÞ ¼ ½yðkÞ; yðk 2 1Þ; . . . ; yðk 2 nÞ
T

ð5Þ

From equations (4) and (5), equation (3) can be rewritten as: y^ ðkÞ ¼ a T ðkÞyðk 2 1Þ þ N ðkÞ

ð6Þ

Then the forecast error is calculated by comparing the actual measured value by the data agents with the AR model predicted value. The prediction error e(k) is defined as the difference between the measured data y(k) and the AR model predicted value _ y ðkÞ, i.e.: _ eðkÞ ¼ yðkÞ 2 y ðkÞ

ð7Þ

Equation (6) shows that the behaviour of the sampled data y(k) has close relationships with the time-series coefficients. During the dynamic forecasting process, the process characteristics change continuously. This causes the sampled data to change as well. As a result, the AR model coefficients, ai s, have been modified to adapt to the change in the sampled data, y(k). Should the error is greater than the perquisite tolerance; the forecasting process stops and the model is re-calibrated or fine-tuned. Otherwise, the process proceeds to update the coefficients of the AR time-series for the next forecasting cycle. The adjustment of the coefficients of the AR time-series is accomplished by an adaptive filter algorithm. In recent years, the adaptive filter has come to play an increasingly important role in the fields of communications (Graupe, 1989), control (Liang and Dornfeld, 1989; Lee et al., 1997), market analysis (Kress and Snyder, 1994) as well as in financial analysis (Haykin, 1984). In the present study, the modification of the AR model coefficient vector a is accomplished by a modified least mean square (MLMS) algorithm (Cheung et al., 2002) which is based on the following equations: aðk þ 1Þ ¼ aðkÞ þ beðkÞyðkÞ and:

A knowledgebased system

ð8Þ

JMTM 17,6

758

eðkÞ ¼ yðkÞ 2 y T ðk 2 1ÞaðkÞ

ð9Þ

where b is the adaptation gain which determines the step size of change of a at each adjustment; y(k) is the measurement vector at kth instance of time which contains the current n 2 1 past signal data samples; e(k) is the prediction error at kth instance of time. For stability or convergence of the MLMS algorithm, the coefficient vector a(k) of the filter approaches the optimum value aopt as the number of iterations k approaches infinity. The MLMS algorithm has two different conditions for convergence, which are convergence in the mean and convergence in the mean square,, respectively, (Haykin, 1984). For the MLMS algorithm, the optimum values for aopt.(0) and yopt.(0) are used instead of the arbitrary values as used in least mean square (LMS) algorithm (Haykin, 1991). They are found by an iteration algorithm which changes each filter parameter, i.e. the order n, the adaptation gain b, the values a(0) and y(0) one by one inside a predetermined iteration range. The set of parameters which gives the minimum value for sum of squared errors (SSE) within the iteration range is selected as the optimum filter parameters. Although the algorithm may search for the sub-optimal parameters, the performance and stability of the filter can be ensured within the iteration ranges. The CBR model starts with chasing the information from past consumption and the information from the enterprise such as the results of material requirement planning, inventory level, demand forecast from the dynamic forecasting model and a series of business rules. Then, a number of advisory recommendations is retrieved and deduced by a forward chaining inference engine. By comparing the standard business rules, the inventory data from the vendors and the company, and the filtered cases, a series of indexes for the recommendations are determined and the corresponding descriptions of the recommendations are retrieved from the reasoning database based on the given indexes. The business rules include spare part demand trend, the control of the demand and lead time formulas, stock usage, planned orders, purchase orders, and vendor performance that reflect unique business cycles. A recommended solution is then selected and adapted by the users as a new decision. The recommendations include availability of spare parts, order quantities, best purchase price, replenishment schedule and terms of references so the right amount of the right materials are available at the right time. The verified case is retained as a new case in the knowledge repository (KR). The description together with the solution of the confirmed case is then stored in the KR for future reuse. For the FSM component, it serves the purposes of optimising the schedule and resources for the field service staff to provide on-site service to the customers. A series of functions are provided for the management staff which include scheduling the service orders, matching the field service staff with customers, optimising the cost of the field service, collecting customer feedback from the customers, reporting and verifying the service records, etc. 3.4 Service automation module The SAM is used to support the automation of the customer service operation off the office hours. Frequently asked questions (FAQs) are provided for the customer to solve their common problems. The FAQs is built based on a rule-based inference engine and the content of FAQs is continuously updated by the customer service staff based on the verified successful cases in the KR. When the customer problems cannot be solved by the FAQs, the customer requests can be reported online anytime and anywhere via the

SAM. Those customer requests are processed by the customer service staff and the customers are notified with the status of their requests via the internet. 3.5 Performance measurement module To ensure the alignment of the quality of service provided by the service staff with the business strategies of the company, a service quality score (S) is determined based on a scoring model. To provide high quality service to the customers, a weighting system is developed to determine the customer privileges. In this weighting system, both field service staff and customers are classified into different classes which are excellence (Class A), average (Class B) and poor (Class C), respectively. Customers of Class A are served by the field service staff of Class A. For the Class B customers, the services are provided by the Class B field service staff or above. The classification of each service staff would be changed based on his/her own performance. By using a scoring model, the score (S) of each field service staff can be determined by a weighted average method which is shown in equation (10): M X W lXl S¼

l¼1 M X

ð10Þ

Wl

l¼1

where M is the number of criteria, Wl is the weighting of the lth criterion, Xl is the normalised value of the lth criterion which should carry value between 0 and 1. The performance is measured by several criteria such as technical skills, servicing period, attitude of the field service staff, etc. The high quality of service score is used as the upper bound score, Supper. When the score of the field service staff is set above the Supper, the performance is graded to be excellent. The baseline analysis of the conventional customer service approach is used to establish the lower bound score, Slower. When the score of field service staff is marked below the Slower, the staff performance is graded to be unsatisfactory. Further action should be made to investigate the problems and those field service staff may need to be re-trained for maintaining high quality of service. Hence, the level of staff performance can be evaluated and monitored by using this scoring model. 4. Case study and discussion Intra-Tech Mechatronics Limited (ITM) is a semiconductor equipment manufacturer based in Hong Kong and dedicated to the supply of equipment for the back end processing of semiconductor devices in chip-on-board (COB) applications. Apart from the normal applications, such as calculator, watches and clocks, ITM equipment are also used for packaging of multiple chip optical communication devices, smart cards and large multiple colour LED display panels. These high-tech applications are required features of equipment with highly accurate placement and deep access capabilities. The mission of the company is to provide the latest semiconductor technology equipment for the customers to manufacture the best quality products all over the world. 4.1 Implementation of service portal In the present study, a service portal is built based on the knowledge-based service automation approach as discussed in Section 3. The service portal is mainly used by

A knowledgebased system

759

JMTM 17,6

760

several parties such as customers, managers, field service engineers and customer service staff. The customers can contact the customer service staff through the internet, which is mainly accessed by three parties such as service managers, field service engineers and customer service staff. The detail relationship among different parties is shown in Figure 5. Figure 6 shows a snapshot of user interfaces illustrating the process of handling the customer requests by the customer service staff. In the service portal, customers can request for customer services through Internet. Before placing an order, the customers need to login the system. FAQs are available for them to solve their common problems. When the customers cannot find any solutions through the FAQs, they can fill in the “service record” to place the service orders. The information of the orders is stored automatically in the KR after confirmation. Sometimes, the customers may place their orders by making a phone call to the customer service staff. At that time, the customer service staff can use the service portal to help the customers to place an order. After finishing all the processes, the records can also be retained in the KR.

Internet Place Order Customers Respond to Customer

User Interface

Get

Save

Knowledge repository

Save Check Available Time for

Save Record

Get

Adapt Cases Service Manager

Figure 5. Entity relationship of the service portal

Retain Cases

Inference Engine

Assign Jobs

Get Data

Suggest Cases

Service Engineers

Report to Clerks

Report to Manager after Verifying

Intranet

Customer service staff

The customer service staff is logged in the system to manage customer request.

A knowledgebased system

761

To retrieve customer records and manage customer requests.

The problems are selected and shown in here.

To click “Confirm” to submit the customer request.

Figure 6. A snapshot of the process of handling the customer requests

For the field service management as shown in Figure 7, the service manager assigns the appropriate field service engineer based on the recommendations from the service portal. Qualified field service engineer is recommended by the portal when he is available and his/her past service performance matches at the same or above the level of the customer privileges. For example, Class A or Class B engineers can provide field service for Class B customers while Class A customers can only be served by Class A engineers only. This ensures that the field service engineers have the capability to handle and serve the customer with ensured service quality. When the field service engineer is assigned the field work, he/she can undertake a pre-assessment of the customer’s problems and retrieve the possible solutions from the

JMTM 17,6 To login as a manager for job assigning and case validation.

762

New customer requests are shown in here.

To assign a suitable engineer to provide on-site services at a certain time.

A frequently used case is validated as a new case/problem for further reuses.

A new problem is shown in the category of “Others” while a case is validated.

Figure 7. A snapshot of the field service management

service portal. Past similar cases are suggested to them based on the KR and KBS modules. The KR of the service portal is composed of case library, databases, business rules and dynamic data for supporting the service logistics. The content of the case and possible solutions is shown in Table I while the taxonomy of the case representation is shown in Figure 8. The KR is built by a MS SQL2000 server system, which provides a scalable and easy-to-integrate database environment for enterprise application integration (EAI). The information is stored in the KR for subsequent decision support. As shown in Figure 9, the case retrieval process starts with entering the customer request and finishes when the best matched cases are identified. When the information of the customer request is entered, the irrelevant information is filtered by using decision tree analysis as shown in Figure 8. Based on the useful information such as operation time, the relevant cases are retrieved. Hence, similarity analysis is used to select the most similar cases based on the nearest neighbour algorithm. Case number

Description

Case indexes

For CRM function

763

Customer information Available time of engineers Grade of engineers Machine number Machine model Operation time Problem type

Description of equipment problem

Solutions

A knowledgebased system

Causes of problems Adapted source case number (indicates the case being adapted) Number of verification Remarks

Table I. Contents of a case in the case library

Equipment

Types of Equipment

Machine Model

Operation Time

Die Bonder

…..

Wire Bonder

…..

BONDA 33

…..

…..

BONDA 103

9500 Hours

Epoxy Dispenser

…..

8000 Hours

…..

Figure 8. Decision tree analysis for filtration

JMTM 17,6 To login and record the maintenance results by a field service staff.

764 The status is “Waiting” while a new job is assigned.

Before providing on-site services, the relevant case is adapted by the field service staff.

The status is changed from “Waiting” to “In Progress” if the case is adapted and well prepared.

The required information is recorded in the service report.

The status is “Verifying” if finishing the on-site service and waiting to verify by a customer service staff. Figure 9. A snapshot of the case handling process

After determining the similarity, the similar previous cases are selected and suggestions are deduced by an inference engine. With the use of forward chaining inference engine, the suggestions such as the causes and recommended solutions for troubleshooting the customer problems are provided for the field service engineer to reuse. The suggested cases are adapted and confirmed by the field service engineers. The associated SPL can be accomplished via the SPL module in the service portal. The demand of the spare parts varies due to a number of factors such as the normal maintenance cycle, seasonal factors, failure patterns, etc. In order to avoid any interruptions of field service due to the unavailability of the spare parts, the inventory of the spare parts are monitored and active replenished (Figure 10) based on the demand forecast determined by the dynamic forecasting model and the recommendations derived from the inventory information and maintenance information by the CBR model. As shown in Figures 11 and 12, the field service engineer can check the inventory level and make use of the service portal for reserving the spare parts. After providing onsite service, the engineers are required to submit a report to the service portal, which will be verified by the customer service staff. In the verification process, the customer service staff may contact the customers for ensuring the accuracy of the information (Figure 13). If some problems are found, corrective actions would be taken. Hence, the new knowledge can be retained in the service portal for future reuse. If the information is different from the previous cases, it would be retained in the portal as a new case. On the other hand, the customer also provides the feedback to evaluate the performance of field service engineer. 4.2 Evaluation of performance and benefits The capability and performance of the service portal are evaluated through two case studies conducted for different service orders. 4.2.1 Case 1: software system upgrade service. In the company, only a few engineers possess the necessary skills to update the software and they are mostly required to work with cross functional teams. As a result, it is time consuming to communicate with each department in order to check the time schedule and job lists. The average total time used to complete the service is reduced from 6 to 5 days. It is interesting to note that the KBSAS approach can achieve a 16.67 per cent saving of the time for handling the service request for the customer. 4.2.2 Case 2: service for storage system replacement. For the conventional approach of storage system replacement, the average total time used to complete the service is around 10.5 days. In the company, no forecasting system is provided to predict demands for spare parts used to provide on-site services. In most cases, the spare parts (e.g. hard disks, mounting assembly, etc.) are easily stocked out. This wastes plenty of time to replenish the materials from a vendor. After the implementation of the KBSAS, more accurate inventory records are kept and this reduces the lead-time of acquiring spare parts for service logistics. As a result, the total time taken for completing the service is reduced from 10.5 to 3.5 days. There is a 66.7 per cent saving of time to handle the customer request. Apart from the efficiency improvement gained as summarized in Table II, the service portal developed based on KBSAS also provides high quality customer services with fast and efficient customer responses, as well as assists the company in acquiring the experience and tacit knowledge in performing customer and field services.

A knowledgebased system

765

JMTM 17,6

To login and request spare parts by a purchaser.

766

To press “Spare Part Advice” button to obtain suggestions.

The quantity forecast, similarity and remarks are given as recommendations.

The case is adapted and revised if needed.

A new spare part request is retained in a knowledge repository.

Figure 10. A snapshot of the spare-part replenishment process

A knowledgebased system

767 To login and select a part for checking inventory level.

The inventory level is shown by plotting the quantity against time.

The records of stocking-in and stocking-out are shown.

Figure 11. Inventory monitoring for spare parts

JMTM 17,6

To login as a field service engineer and reserve spare parts before providing onsite services.

768

To fill required data such as required quantity, warehouse no, department, etc.

If the reservation is made successfully, the new record will be displayed as a new reservation. Figure 12. Reservation of spare parts for field service

At current stage, the portal yields a number of advantages over conventional service logistics, which include: . streamlining the service logistics process; . provide 24 hours worldwide automatic customer service which is supported by the verified knowledge base established in the date time operations; . drive for continuous improvement of customer service quality; . reduce paper work; . knowledge can be captured in terms of verified cases in the KR; . knowledge can be shared among the staff; and . staff performance can be measured. 5. Conclusions Although the CSM concept has been applied in different sectors of industries such as banks, insurance and telecommunications, the advantages of CSM systems are generally overlooked in manufacturing industries. Service logistics is indispensable to

A knowledgebased system

The customer service staff logs to verify cases.

769 While the status of case is “Verifying”, the button of “Detail” is clicked to continue processes.

To revise the problem or give remarks for corrective action.

After clicking “Confirm”, the status of case is changed from “Verifying” to “Completed”

Figure 13. Verification of the service performance

Case descriptions Software system upgrade service Service for storage system replacement

Conventional approach (days)

Knowledge based-service automation approach (days)

Total reduced time (per cent)

6 10.5

5 3.5

16.67 66.67

Table II. A comparison between the conventional and knowledge-based service automation approaches for handling different service orders

JMTM 17,6

770

achieve effective and efficient CSM. However, the conventional ways of the service logistics are information driven instead of knowledge-driven which are inadequate to meet the current needs. To provide a higher value service and to achieve a better customer satisfaction, a knowledge-enabled infrastructure is much needed for achieving multi-perspective of an enterprise in knowledge acquisition, service automation, service logistics streamlining and hence measurement of the performance of its service processes. By integrating the knowledge-based and AI technologies, a KBSAS is proposed for service logistics. The capability and advantages of the KBSAS are demonstrated through the trial implementation of a prototype service portal in a high-tech semiconductor equipment manufacturer. It is verified that the KBSAS provides high quality customer services with fast and efficient customer responses. It also allows the company to capture the valuable experience and tacit knowledge of the staff in performing customer and field services. This helps to drive the continuous improvement of customer service quality. References Aamodt, A. and Plaza, E. (1994), “Case-based reasoning: foundational issues. Methodological variations, and system approaches”, AI Communications, Vol. 7 No. 1, pp. 39-59. Anderson, K. and Kerr, C. (2002), Customer Relationship Management, McGraw-Hill, New York, NY. Bose, R. (2002), “Customer relationship management: key components for IT success”, Industrial Management & Data Systems, pp. 89-97. Cheung, C.F., Lee, W.B., Lo, V. and Wang, W.M. (2002), “Dynamic forecasting using a modified least mean square algorithm”, Proceedings of the 10th International Manufacturing Conference (IMCC2002), October 11-13, Xiamen, China, 1-238. Cheung, C.F., Lee, W.B., Wang, W.M., Chu, K.F. and To, S. (2003), “A multi-perspective knowledge-based system for customer service management”, Expert Systems with Applications, Vol. 24 No. 4, pp. 457-70. Cheung, C.F., Wang, W.M. and Kwok, S.K. (2005), “Knowledge-based inventory management in production logistics: a multi-agent approach”, Proceedings of The Institute of Mechanical Engineers, Part B, Journal of Engineering Manufacture, Vol. 219 No. 3, pp. 299-308. Feinberg, R. and Kadam, R. (2002), “E-CRM web service attributes as determinants of customer satisfaction with retail web sites”, International Journal of Service Industry Management, Vol. 13 No. 5. Finnie, G. and Sun, Z. (2003), “R5 model for case-based reasoning”, Knowledge-Based Systems, Vol. 16, pp. 59-65. Graupe, D. (1989), Time Series Analysis, Identification and Adaptive Filtering, 2nd ed., Robert E. Krieger Publishing Company, Huntington, NY. Haykin, S. (1984), Introduction to Adaptive Filters, Macmillan, New York, NY. Haykin, S. (1991), Adaptive Filter Theory, 2nd ed., Prentice-Hall, Englewood Cliffs, NJ. Jiang, P.Y., Zhou, G.H. and Liu, Y. (2002), “ASP-driven e-service platform for we-based online manufacturing”, Integrated Manufacturing Systems, Vol. 13 No. 5, p. 318. Kress, G.J. and Snyder, J. (1994), Forecasting and Market Analysis Techniques: A Practical Approach, Quorum Books, Westport, CT. Langer, M., Loidl, S. and Nerb, M. (1999), “Customer service management: towards a management information base for an IP connectivity service”, paper presented at The Fourth IEEE Symposium on Computers and Communications, Red Sea, Egypt, pp. 149-55.

Lee, J. (2003), “E-manufacturing – fundamental, tools, and transformation”, Robotics & Computer-Integrated Manufacturing, Vol. 19 No. 6, pp. 501-7. Lee, W.B., Cheung, C.F., Chiu, W.M. and Chan, L.K. (1997), “Automatic supervision of blanking tool wear using pattern recognition analysis”, International Journal of Machine Tools & Manufacture, Vol. 37 No. 8, pp. 1079-95. Liang, S.Y. and Dornfeld, D.A. (1989), “Tool wear detection using time-series analysis of acoustic emission”, Journal of Engineering for Industry, Vol. 111, pp. 199-205. Simoudis, E. (1992), “Using case-based reasoning for customer technical support”, IEEE Expert, Vol. 7 No. 5, pp. 7-13. Watson, I. and Marir, F. (1994), “Case-based reasoning: a review”, The Knowledge Engineering Review, Vol. 9 No. 4. Widrow, B. and Hoff, M.E. Jr (1960), “Adaptive switching circuits”, IRE WESCON Conv. Rec., Pt 4, pp. 96-104. Zhang, Y.F., Jiang, P.Y. and Zhou, G.H. (2003), “GA-driven part e-manufacturing scheduling via an online e-service platform”, Integrated Manufacturing Systems, Vol. 14 No. 7, p. 585. About the authors C.F. Cheung is an Associate Professor in the Department of Industrial and Systems Engineering of The Hong Kong Polytechnic University and an Adjunct Professor of the Harbin Institute of Technology Shenzhen Graduate School. His research interests include logistics systems, e-business, artificial intelligence, knowledge management, and precision engineering. C.F. Cheung is the corresponding author and can be contacted at: [email protected] Y.L. Chan is an MPhil student of the Department of Industrial and Systems Engineering of The Hong Kong Polytechnic University. She received her BEng degree in Manufacturing Engineering at the same University. Her research interests include supply chain management, logistics systems and knowledge management. S.K. Kwok is a Lecturer of the Department of Industrial and Systems Engineering of The Hong Kong Polytechnic University. His research interests include mobile commerce, radio frequency identification technology, and logistics systems. W.B. Lee is the Chair Professor and Head of the Department of Industrial and Systems Engineering of The Hong Kong Polytechnic University. He is also the Directors of the Microsoft Enterprise Systems Centre and the Advanced Manufacturing Technology Research Centre of The Hong Kong Polytechnic University. His research interests include advanced manufacturing technology, manufacturing strategy, and knowledge management. He also acts as the editorial board member in a number of international journals. W.M. Wang is a PhD student of the Department of Industrial and Systems Engineering of The Hong Kong Polytechnic University. He received his BEng degree in Manufacturing Engineering at the same University. His research interests include electronic business systems and knowledge engineering.

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

A knowledgebased system

771

The current issue and full text archive of this journal is available at www.emeraldinsight.com/1741-038X.htm

JMTM 17,6

772

The applicability of a multi-attribute classification framework in the healthcare industry Konstantinos Danas

Received September 2005 Revised January 2006 Accepted March 2006

The American College of Thessaloniki, Thessaloniki, Greece

Abdul Roudsari Centre for Health Informatics, City University, London, UK, and

Panayiotis H. Ketikidis City Liberal Studies, an Affiliated Institution of the University of Sheffield, Thessaloniki, Greece Abstract Purpose – To introduce the applicability of the Ned-MASTA classification method for medicines within the environment of a hospital pharmacy and the virtual pharmacy inventory system that forms a virtual pharmacy inventory of hospitals within the same geographical region providing the infrastructure for the cooperation of hospital pharmacies in order to improve the efficiency of their operations. Design/methodology/approach – A survey that was conducted in Greek hospitals identified the inefficiencies of their logistics systems that are similar to inefficiencies identified through surveys in hospitals worldwide. It was considered vital and necessary to investigate the solutions that are provided in other industries facing similar problems. The case of spare parts inventory for production machines was found to present similarities with the management of medicine stock within the hospital pharmacy. The approach that was followed for the case of spare parts was modified and included in the system that forms a virtual hospital pharmacy inventory; this made the approach applicable in the hospital environment and further improved the efficiency of the use of hospital pharmacy resources. Findings – The MASTA classification approach was modified to fit in the operation of the hospital pharmacy and a system was constructed to form the virtual pharmacy inventory. The applicability of the system is demonstrated through an application scenario. Research limitations/implications – The system is in the form of a prototype under evaluation. It has not been applied yet thus results that are based on actual applications are not presented. Practical implications – It demonstrates the idea of a solution to the inefficiencies of the hospital pharmacy and sets the ground for discussing the proposed solution. Originality/value – This study introduces a new approach to the problems and inefficiencies of the hospital pharmacy management. Keywords Health services, Supply chain management, Classification schemes, Greece Paper type Research paper Journal of Manufacturing Technology Management Vol. 17 No. 6, 2006 pp. 772-785 q Emerald Group Publishing Limited 1741-038X DOI 10.1108/17410380610678792

1. Introduction and background of the problem domain The healthcare industry is considered the world’s largest industry with a budget of more than three trillion US$ and tens of millions of people under its employment

(Kirsch, 2002). In the UK, 10 per cent of the working population bases their income on the National Health Service (NHS) either as employees or as suppliers (Brennan, 2005). The costs related with the healthcare industry are dramatically increasing in the last few years, from 3.9 per cent of GDP to 7.6 per cent, according to the figures of the Organisation of Economic Cooperation and Development (OECD) and only within the USA, healthcare expenditure has reached 13.6 per cent causing concern to governments who try to identify ways to reduce healthcare expenditure, and in this way, increase the efficiency of the industry and improve the safety and quality of patient care (Jarret, 1998; Whitson, 1997; Wilson et al., 1992; Faldala and Wickamasinghe, 2004; Okoroh et al., 2001). There are many factors affecting these cost increases like the ageing population, the technological, medical and biological advances that have taken place within the last decades and are due to carry on in the future and the inefficient use of the industry’s resources. Errors or omissions during patient treatment leading to adverse events and medication errors add four billion pounds in the UK budget and more than five million dollars in the USA budget (Brennan, 2005). Many of the accrued inefficiencies are directly related to poor information flow within the healthcare environment. Even though information technology (IT) has been adopted and has increased the efficiency in other industries, the healthcare industry has not yet benefited from the introduction of IT tools. Regarding patient safety and the prevention of medication errors, decision support systems and computerised physician order entry systems (CPOEs) based on the electronic patient record have been successfully implemented. Such systems are able to check patient prescriptions, calculate drug dosages for specific patients and also check for allergies and other drug reactions that could prove harmful to patients. As these systems are able to calculate the correct dosage of the correct drug that should be given to patients, a logistics system is necessary to ensure that this drug is in stock and in the proper condition when required. Governments then proceed to issuing the essential directives for optimising the operations of the healthcare industry and at the same time for improving the service, quality and safety to patients. These directives are based on the science of logistics that has provided models and concepts already applied in other industries, making them more effective and more profitable. They include directions and standards to be adopted by the industry in order to build the proper infrastructure for the facilitation of information exchange. Such directives include the EHCR initiated by the Clinton administration in the USA and also adopted in Canada, PECC in Australia and others in the EU. All these directives aim at increasing patient safety and expenditure efficiency by considering models that are based on the science of logistics and on contemporary information systems like enterprise resource planning (ERP) systems, which is based on information integration within an organisation and between this organisation and its partners. In order to optimise the hospital logistics system and the healthcare supply chain, successful practices of other industries like for example the just in time (JIT) model of supplies that aims to reduce inventories and to improve quality and services to the customers reducing at the same time the cost of logistics activities, are under consideration for adoption. The healthcare industry has been reluctant in the adoption of JIT since JIT is based on accurate forecasting and in the healthcare industry the patient mix cannot be accurately predicted (Jarrett, 1998). Furthermore, hospitals deal on a daily basis with emergency situations that require

Applicability of a multi-attribute classification 773

JMTM 17,6

774

immediate actions in order to save lives. Based on JIT, several models evolved like the stockless inventory system (Wilson and Cunningham, 1992) and the hybrid stockless (Rivard-Royer et al., 2002) that have been tested and proven valuable. These models are based on the use of ERP systems including the use of bar codes for product identification and electronic data interchange (EDI) for communication with their partners. The hospital pharmacy has not been considered yet for a JIT strategy of supplies since emergency cases, where the hospital must provide the proper medication to patients, make forecasting inapplicable. Therefore, the hospital pharmacy holds stocks of drugs in order to be able to respond to emergency situations, building large and difficult to manage inventories of drugs. In a survey that was conducted in public and private hospitals in the area of Northern Greece (Danas and Ketikidis, 2000), the inefficiencies of the hospital pharmacy inventory management were identified. Such inefficiencies included stock out incidents, which endangered patient safety and life, incidents of expired drugs that had to be thrown away as ineffective. Pharmacists tend to spend most of their time in inventory management and not on patient treatment and safety. On the other hand, some drugs must be inventoried in case they are needed and this forms part of the problem. At this point, it is important to investigate solutions that have been developed in other industries facing similar problems with the hospital pharmacy where forecasting becomes inapplicable. The case of spare part inventories for production machines in large industrial plants is similar to the case of the hospital pharmacy. Spare parts inventories support the service of production machines when such machines break down. Down times must be minimised, as the cost of a down time may be enormous. Spare parts inventories are dominated by uncertainty; they store very expensive spare parts that may never be used but in case that a spare part is needed, they ensure the minimisation of the machine down time and continuous plant production. Contemporary supply chain management techniques greatly depend on continuous production since they are based in JIT to lower inventories and ensure the smooth flow of products from production to consumption. An approach to a solution of the problems of the hospital pharmacy could be based on a classification technique that has already been adopted in the case of spare parts inventories and has been proved beneficial. As public hospitals do not compete with each other, they could form a network in order to be able to exchange drugs and eventually form a virtual pharmacy over a certain geographic region (Danas et al., 2002a). In this way hospitals would be able to support each other in cases of emergency situations thus reducing the dead stock they keep for such cases. They could also exchange drugs that are about to expire – before their expiration date – but might be needed by other hospitals further improving in this way their operation efficiency. By reducing the fear of an emergency stock out, an approach based on JIT could be considered for adoption as so far, resistance to the adoption of JIT in healthcare is based on such cases. The aim of this study is to introduce a drug classification system that could enable a JIT supply strategy to be adopted by a hospital pharmacy. This classification system is based on the virtual pharmacy inventory system (VPIS) that has been previously introduced and forms a virtual inventory of drugs, virtually connecting hospital pharmacy inventories in the same geographical area (Danas et al., 2002a, b). Based on

VPIS, hospitals can support each other in cases of emergencies and exchange drugs that are about to expire. The study is organised in two parts; the first part presents the case of spare parts inventories and identifies the similarities between spare part inventories and the hospital pharmacy. The second part presents the classification technique for drugs, its role within the VPIS and an application scenario to demonstrate the applicability of the hospital pharmacy management model. The study concludes with discussion, conclusions and further thoughts on this new approach. 2. Review of spare parts inventory models Supply chain management is considered as a strategy aiming at the efficient flow of products from the point of production to the point of consumption. This is possible by reducing inventories, using more frequent product deliveries, by improving the quality of service to customers, by reducing the overall logistics costs and improving the efficiency at the same time. Supply chain management is based on accurate planning, information flow and partner coordination. The formation of supply chains includes industrial organisations that have to produce and forward their products to the next node of the supply chain that is considered as their customer. Within the supply chain, it is very important that each participant delivers the products to the next node at the time that they are needed. Industrial organisations must be able to support continuous production maximising the utilisation of their production machines. In case that a production machine fails and needs repairing, servicing and spare parts, the downtime must be minimised, as downtime costs are dramatic for the whole supply chain. The inventories of spare parts are then essential to ensure the minimisation of downtimes and related costs. Such inventories are expensive; therefore spare parts inventory management becomes crucial and any developments in the management of such inventories could reduce investment and improve performance. The majority of the literature review on stock control focuses on distribution networks and forecasting techniques that could be reliable and used in order to predict demand. In spare parts inventories, demand is not accurately predicted and at the same time the stock is very expensive. ERP systems that have the functionality to accurately calculate safety stocks and produce forecasts cannot be used to manage slow moving material with unpredictable demand such as spare parts (Razi and Tarn, 2003). The questions concerning the situation under investigation are the following: . which spare parts to stock; . where are these spare parts stocked; and . how many units to stock (Botter and Fortuin, 2000). According to Braglia et al. (2004), there are two approaches to answer the above questions, mathematical modelling and classification. Mathematical models are too complex, abstract or simplified; therefore, their utility is limited to the maintenance managers. Classification systems that are based on ABC methods concentrate mainly on price and quantity issues limiting their ability to focus on other attributes of each part. Given that each part has different characteristics, fulfils different needs, has different size, different obsolescence characteristics, different price, classical classification approaches are limited. To overcome this limitation, multi attribute classification methods based on the management of multiple factors and supported by expert information systems are introduced (Braglia et al., 2004).

Applicability of a multi-attribute classification 775

JMTM 17,6

776

3. Spare parts classification model description Some classification approaches available in the literature are based on the classical ABC analysis. The limitations of this approach as mentioned earlier in the study are overcome by multi-attribute classification models able to manage multiple criteria and factors, sometime conflicting with each other. One such approach is the one that has been put forward by Braglia et al. (2004) namely the multi attribute spare tree analysis (MASTA). The MASTA approach has been considered for the purpose of this study to identify the similarities of spare parts inventories and hospital pharmacy inventories. The MASTA approach is used in spare parts inventory management to answer the above-mentioned questions of spare parts inventory management. Braglia et al. demonstrated MASTA by using it to determine the inventory policy for a large paper mill, thus the analysis that follows concerns the specific case study. It is based on two consecutive steps and its results concern the formation of a matrix that identifies the inventory policy regarding each spare part stock and supply strategy. The first step concerns the criticality analysis based on logic trees. In the case study it is recognised that each spare part is checked to identify its criticality for four major categories, namely the criticality to the plant, the supply characteristics, the inventory problems and usage rate. For each major category, there are further characteristics that are examined. The criticality of each of these characteristics can be critical, important and desirable. The whole process starts examining each major category in the order presented (i.e. plant criticality first) and within each major category its characteristics are evaluated to be critical, important or desirable. At the end, the whole category is evaluated. According to the evaluation of each category a different path is followed. At the end of the whole process each part is checked against all categories and sub characteristics and is assigned a class. According to the class of each spare part, the inventory policy matrix indicates the inventory policy strategy. The model was evaluated using simulation of the past five years of machine failures (Braglia et al., 2004). 4. Methodology A survey that has been conducted in public and private hospitals of Northern Greece identified much inefficiency of their logistics systems (Danas and Ketikidis, 2000). These inefficiencies are similar to others that were identified in the healthcare industry worldwide. It is obvious that the solutions and directives that were provided, as mentioned above, should also be adopted in the Greek healthcare industry. Such solutions provide the directives and the proper infrastructure that integrated information systems (ERP) for hospitals should be based on. These approaches aim at the facilitation of information flow inside the hospital and between hospitals and their partners like suppliers, insurance organisations, etc. Other approaches are based on the reduction of hospital inventories, shifting the management of stock to their suppliers. The situation could be improved if hospitals in the same geographical area coordinate with each other, in which case, the infrastructure for such coordination must be added to the above-mentioned initiatives. Having identified the problems and inefficiencies of the current situation a systems analysis and design methodology must be followed in order to produce the new system. The methodology that has been selected for this study is the abstraction of the Zachman framework put forward by (Whitten et al., 2001).

In order to investigate the applicability of MASTA in healthcare and its modification for the hospital pharmacy, doctors and pharmacists were interviewed. MASTA was presented to three doctors and then with an open discussion the categories and attributes of the healthcare industry were identified. These categories were then presented to hospital pharmacists who had the opportunity to remark and comment on the categories and attributes. The practitioners’ help was considered very important for the modification of MASTA for the hospital pharmacy. Once a prototype of the new system, including the classification framework, is produced, a structured evaluation methodology will be carried out to surface the qualities and usability of the system.

Applicability of a multi-attribute classification 777

5. Drug classification model description The idea is to follow the structure of MASTA and test each drug separately in order to identify the proper stock and inventory strategy for it. Structuring the classification model, the major categories need to be identified and for each one of them the characteristics also need to be attributed. At the second step the class of each drug will determine the inventory and stock strategy for it in order to optimise the hospital pharmacy inventory. The expert advice of doctors and pharmacists is essential during this phase in order to define the categories and characteristics of drugs. The aim of the healthcare industry is to provide quality and safe treatment to patients. Given that the focus of attention is the patient, the role of the supply chain is to be able to provide the correct drug at the time that the patient needs it. The model for drug classification would be called Med-MASTA (Figures 1-4). The structure of Med-MASTA is the following: (1) Patient treatment criticality; . danger of loss of life (1: critical, 2: important, 3: not important); . quality of treatment; . replacement with other treatment; and at the end a weighted average describes the level of patient treatment criticality (1: critical, 2: important, 3: not important).

Patient Criticality

Sub Tree A

Sub Tree B

Sub Tree C

Figure 1. The sub-trees

JMTM 17,6

778

(2) Supply characteristics; . lead time; . number of potential suppliers; and . replacement (3) Inventory problems; and . price; . space required; . special inventory condition; and . expiry date. (4) Usage rate . over stocking; and . frequency of use. The same criticality levels as the patient treatment criticality are applied to all characteristics and all categories that are examined.

Critical

Critical

Supply Characteristics

Patient Treatment Criticality

Not Important

Important Inventory Not Important

Not Important

Inventory Problems Important Critical

Critical

Usage Rate

Important or Not Important

Figure 2. The sub-tree A

A

B

B

A

A

Not Important

B

Important

Inventory

C

Important

Critical

B

Not Important

Inventory

Use

A

Critical

Not Important

Not Important

Critical

Important

Supply

Important

Patient Criticality

Critical

C

Not Important

Use

Inventory

B

Critical

B

Applicability of a multi-attribute classification 779

Figure 3. The sub-tree B

B

Critical

Figure 4. The sub-tree C Inventory

B

Use

C

Not Important

Supply

B

Not Important

Not Important

C

Critical

Critical

Inventory

Use

780

Critical

Critical

Not Important

Not Important

Patient Criticality

D

Not Important

JMTM 17,6

The drugs would be classified in the following categories: . A: very important. The very important drugs should be stocked in each clinic that uses them and at every hospital pharmacy as safety stock. . B: important. The important drugs should be stocked in each clinic that uses them but the safety stock is distributed among hospitals in the same geographical area, and the virtual hospital inventory system (VPIS) should be able to manage the safety stock. . C: less important. The less important drugs should be stocked only in each clinic that uses them, making each clinic responsible for the stock management. . D: not important. Not important drugs should be supplied in a JIT basis in each clinic that requires them. The above classification system and inventory strategies are based on the VPIS model to support each clinic and hospital that may require a drug and they do not currently have it in stock. VPIS model is able to search and locate required drugs that are stocked within the hospitals that form the network. 6. Description of the system VPIS is considered as a complementary system to the hospital ERP system that provides a platform for the cooperation among hospitals in the same geographical area (Danas et al., 2002b) and a decision support system for the effective management of the hospital pharmacy. The hospital pharmacist will be the main user of the system but the system could also be used by doctors and head nurses responsible to manage their clinic pharmacy stock. The pharmacists will be able to check drugs that could be in offer to the network by other hospitals before completing their daily supply order forms so that drugs would be circulated within the hospital network and not wasted if they are close to expiring. This aspect of the system is particularly helpful for antibiotic drugs that are stored in the hospital pharmacy in case of an Anthrax attack. Such antibiotics that are also used for other purposes are not wasted and distributed to the other hospitals that may need them. Doctors would be able to locate a drug they need for an emergency situation that may not be in stock within their clinic or hospital pharmacy. It is important to mention that the observation of such an incident provided the central idea for the cooperation between hospitals. During one hospital pharmacy visit, a drug was required to cure a frozen leg incident. The pharmacist made an effort to locate the drug for the above treatment wasting more than three hours on the phone with other pharmacists. Head nurses and pharmacists could also use the system to replace drugs that they find below the safety stock in their hospital and are overstocked in other hospitals. As part of the VPIS, Med-MASTA classifies the drugs that are stored in the hospital pharmacy according to parameters that doctors and pharmacist provide to the method characteristics. The drug classification helps the hospital pharmacist to decide what inventory strategy to use for each drug separately (Figure 5). Based on this classification, the hospital pharmacy inventory could be dramatically reduced, since only a portion of drugs would be kept as safety stock in each hospital pharmacy inventory. The system would be based on a group of static agents being responsible to integrate the VPIS with the hospital database and then form the database of VPIS.

Applicability of a multi-attribute classification 781

JMTM 17,6

7. Virtual pharmacy inventory system technical description As mentioned earlier, VPIS is based on a group of agents included in a batch file that should run at specific time periods updating the system’s databases interfacing VPIS with the hospital’s ERP system (Figure 6). The task of these agents is to interrogate the HOSPITAL PHARMACY SAFETY STOCK

782

Figure 5. The inventory strategy matrix

A B C D

VIRTUAL PHARMACY SAFETY STOCK

X

CLINIC SAFETY STOCK

JIT NO SAFETY STOCK

X X X

X

X

VPIS Description

Batch File VPS Clasif MedMASTA Offer CentralOffer

VPIS Database Medicine Clasif MedMASTA Offerings Central Offer

Hospital db Offer

Read

Hospital Information System

Create

Get class

Read

VPS Medicine Table

Create new table Offerings Table Read

Med-MASTA VPIS Control (Vcontrol)

Central Offer

Update Clasif

Create newtable

Med-MASTA Read Clasif

Med-MASTA parameters

Central Offerings Table

Check stock data New Order

Locate Emergency VPIS Control (Vcontrol)

Figure 6. The VPIS description

Users use the system through the Vcontrol program

Hospital Network

hospital’s database and then produce the database tables that are used by the system. Once the batch is executed, the drugs that should be circulated within the hospital network are stored in a database table once a specific set of rules has been applied to the drugs that this hospital has in stock. A table consisting of the offers of all the hospitals in the network is then produced. The use of the system is achieved by the control program of VPIS that is installed in all hospitals in the network providing the user interface. Users from each hospital would be able to use the VPIS through this control program. 8. An application scenario The new JIT system model is based on the reduction of the central pharmacy inventory of the participating hospitals. The Med-MASTA classification technique classifies the drugs producing the supply strategy matrix as described earlier. This way the safety stock of class “A” drugs will remain as it is. In other words this safety stock will be stored in the central hospital pharmacy and the clinics that use the specific drug. Such drugs are considered very important for the hospital pharmacy operation. Drugs that are classified as class “B” are stored in the clinic that uses them and the safety stock is spread among the hospitals in the network. If for example the safety stock of a class “B” drug is 1,000 units and the hospitals participating in the network are ten then each hospital will store 100 units in its central pharmacy of this specific drug stock. The safety stock of this drug would be located from one hospital to another using the VPIS. Pharmacists that require a portion of the stock would be able to locate it in another hospital and then have it transferred in case of an emergency situation. Class “C” drugs are only stored to the clinic that uses them, in other words the safety stock is spread among the clinics within the hospital. Again, in a case that such a drug is needed and is not in the clinic stock, the drug would be located using VPIS. Finally drugs of class “D” are only ordered on a JIT basis directly to the clinic that requires them. For such drugs a safety stock does not exist. The head nurse will transfer the quantities required for the next day and the pharmacist is then responsible to proceed with this order. Using the results of the classification, the central hospital pharmacy will end up storing and managing the safety stock of the drugs of class “A” and a small portion of the safety stock of the drugs of class “B”. On the other hand, the hospital pharmacists would have the responsibility to process more orders from the clinics to the suppliers but modern ERP systems that hospitals use are eligible to assist. The pharmacist is also going to use VPIS in order to check the offerings of other hospitals whenever a new order for drug supplies must be edited (A later version of the system could be able to perform this task automatically, as it is a matter of further integration with the hospital ERP system and actual implementation of VPIS.). This way drugs that are about to expire will not be wasted or returned to the suppliers. This will lead to further cost reduction and more efficient use of pharmacy resources. The pharmacist would also use VPIS to check the safety stock of drugs and if one is found under the safety stock level then VPIS can output the amount of overstock of that specific drug in other hospitals within the network. This way, the pharmacist would be able to get the required amount of that drug immediately and then order it from a supplier, eliminating the possibility of a stock out. In case of an emergency situation where a doctor or a pharmacist cannot support a specific treatment due to a drug stock out, VPIS is able to check all other hospitals in

Applicability of a multi-attribute classification 783

JMTM 17,6

784

the network and locate the drug that is urgently needed informing the user about the location and the quantity in which this drug is stored. 9. Summary and conclusions The logistics science models and concepts that have successfully been implemented in other industries must also be considered for the healthcare industry in order to make it more efficient. The models that should be adopted must provide similarities when considered from one industry to another. Such a case is the MASTA classification framework that has successfully been implemented in the case of spare parts inventories for the production machines of industrial plants. The similarities of such inventories to the inventory of the hospital pharmacy form the basis for the development and implementation of this classification approach to classify each drug that is stored in the hospital pharmacy considering multiple characteristics. The experience of doctors and pharmacists was utilised to develop a classification framework for application in the hospital pharmacy. This classification framework, called Med-MASTA, is implemented as a part of VPIS, a system that forms a virtual pharmacy inventory of hospital pharmacies within a geographical region. As part of VPIS, Med-MASTA forms the supply strategy matrix and using VPIS the hospital pharmacists could be able to manage their pharmacy inventories more efficiently. The limitation of the system is that it depends on the hospital ERP system in order to provide accurate and reliable information regarding the drugs that are entered in its databases. Even though the results of the Med-MASTA could be used independently and provide an indication to form a supply strategy for each drug separately, VPIS is necessary in order to form the infrastructure for the actual use of the supply strategy, since the stock of many drugs would be spread in the virtual inventory, making VPIS an essential tool to locate it. Future research includes a pilot study, focused on the operation of a specific clinic within a hospital. The system must be parameterised with the aid of a doctor and a hospital pharmacist. Once the parameters of the Med-MASTA are entered to the system, the results will indicate how to form the supply matrix and early indications about the benefits of its use will be identified. Another direction for future research will be the integration of VPIS with a CPOE. Using a CPOE, doctors will enter the patient prescriptions that are checked for patient safety reasons. Once the prescription is verified, the use of VPIS will help in the location of the required drugs. References Braglia, M., Grassi, A. and Montanari, R. (2004), “Multi-attribute classification method for spare parts inventory management”, Journal of Quality in Maintenance Engineering, Vol. 10 No. 1, pp. 55-65. Brennan, S. (2005), The NHS IT Project: The Biggest Computer Program in the World Ever!, Radcliffe Publishing Ltd, Abington, MA. Botter, R. and Fortuin, L. (2000), “Stocking strategy for service parts – a case study”, International Journal of Operations & Production Management., Vol. 20 No. 6, pp. 656-74. Danas, K. and Ketikidis, P. (2000), “E-logistics: a framework for the hospital pharmacy”, Eurolog 2000 Conference Proceedings, pp. 114-21.

Danas, K., Roudsari, A.V. and Ketikidis, P. (2002a), “A virtual hospital pharmacy inventory: an approach to support unexpected demand”, International Journal of Medical Marketing, Vol. 2 No. 2, pp. 125-9. Danas, K., Roudsari, A.V. and Ketikidis, P. (2002b), “The VPIS system: a new approach to healthcare logistics”, Intelligent e-Health Applications in Medicine. Special Edition Proceedings of the EUNITE – 2002 Symposium. Faldalla, A. and Wickramasinghe, N. (2004), “An integrative framework for HIPAA-compliant I *IQ healthcare information systems”, International Journal of Healthcare Quality Assurance, Vol. 17 No. 2, pp. 65-74. Jarrett, G.P. (1998), “Logistics in the health care industry”, International Journal of Physical Distribution & Logistics Management, Vol. 28 No. 9, pp. 741-72. Kirsch, G. (2002), “The business of e-health”, International Journal of Medical Marketing, Vol. 2 No. 2, pp. 106-10. Okoroh, M.I., Gombera, P.P., Evison, J. and Wagstaff, M. (2001), “Adding value to the healthcare sector-a facilities management partnering arrangement case study”, Facilities, Vol. 19 Nos 3/4, pp. 157-63. Razi, A. and Tarn, M. (2003), “An applied model for improving inventory management in ERP systems”, Logistics Information Management, Vol. 16 No. 2, pp. 114-24. Rivard-Royer, H., Landry, S. and Beaulieu, M. (2002), “Hybrid stockless: a case study. Lessons for health-care supply chain integration”, International Journal of Operations & Production Management, Vol. 22 No. 4, pp. 412-24. Whitson, D. (1997), “Applying just in time systems in health care”, IIE Solutions, Vol. 29 No. 8, pp. 33-8. Whitten, J.L., Bentley, L.D. and Dittman, K.C. (2001), Systems Analysis and Design Methods, 5th ed., McGraw-Hill, New York, NY. Wilson, J.W. and Cunningham, W.A. (1992), “Stockless inventory systems for the healthcare provider: three successful applications”, Journal of Healthcare Marketing, Vol. 12 No. 2, p. 39. Corresponding author Konstantinos Danas can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

Applicability of a multi-attribute classification 785

The current issue and full text archive of this journal is available at www.emeraldinsight.com/1741-038X.htm

JMTM 17,6

786 Received September 2005 Revised January 2006 Accepted February 2006

Systems and application development for portable maintenance aid (PMA) – a performance perspective Tim S. Leung, Ka Wing Lee and Walter W.C. Chung Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Kowloon, Hong Kong, People’s Republic of China Abstract Purpose – This paper aims to explore the theory of technology adoption using a system-application approach to facilitate PMA innovation span across multiple organisations to enhance overall production network effectiveness. Design/methodology/approach – The survey on technology available and forecast future trends, design a workable prototype for test in real-world airline operation under a three-layer analysis model. Data is taken to validate the concept and correlate performance factors in PMA adoption. Findings – The findings highlight greater operational efficiency and higher process value maintenance to be achieved through communication paths. The value process-to-shop concept is validated, next to value network. Practical implications – This research supports maintenance business, especially with high contents of information and knowledge driven tasks for outsource service providers, can make use of the wireless system adoption rules for assets and maintenance performance management. Originality/value – The paper presents organisations with a tool such that portable maintenance device can be identified as contribution in different levels, focus and functional effects. Within a critical process, airlines can better manage their dynamic operation process for higher revenue generation. Keywords Software tools, Performance measures, Productive maintenance, Information systems, Aircraft Paper type Research paper

1. Introduction This paper introduces a “systems-application” approach for mobile information system development for commercial aircraft maintenance and engineering (M&E) business in pursuance of the economic advantage of portable maintenance aid (PMA) systems for the production network players. M&E industry consists of the airlines’ engineering divisions, maintenance service providers and multiple tiers of suppliers.

Journal of Manufacturing Technology Management Vol. 17 No. 6, 2006 pp. 786-805 q Emerald Group Publishing Limited 1741-038X DOI 10.1108/17410380610678800

Portable Maintenance Aid (PMA) is generic name for software tool provided by the original equipment manufacturers (OEM) which converts all the paper documents relating to the equipment maintenance and repair to digital formats. Also, PMA programmes enable maintenance information transmission through general portable computing devices. The authors acknowledge the support offered by Central Grant of Project Codes RG45H and RGAE of the Hong Kong Polytechnic University and airlines and M&E users to offer their valuable feedback for Crystal1 prototype evaluation and specification creation.

This industry is characterised by intensive technical information managed tasks, dispersed operations in global scale and fast response time to the minutes. It is conceived suitable to snap on the productivity edges offered by mobile technology. If PMA is employed intelligently as exemplified by So (2004), Lee (2003) and Mennecke (2003) it should drive significant cost saving and effective process for maintenance, and better maintenance means better aircraft utilisation for revenue generation. However, comparing airliner’s PMA with mobile successes in other business applications, e.g. speed post, warehouse, estate and entertainment management as their crucial process as depicted through case studies by Davies (2000), Nunn (2001), Quah (2002), Pahl (2003), McBratney (2003) and Zografos (2002), mobile data support for airline M&E is still rudimentary. Many attempts as reported by Fletcher (2002), Ogaji (2002), Rao (2003) and Greenough (2001) to develop maintenance PMA are reported but general adoptions to large-scale M&E operation is uncommon. We incline to explore a roadmap forward to deploy mobile technology in this industry. 1.1 Stakeholders One problem identified for mobile system development according to Lu et al. (2005) is the division of interest between the system stakeholders and the customers. For research purpose the managers in commercial airliners are regarded as stakeholder since on-time despatch is the important deliverable for contracted or in-house maintenance service. Chan (2005) revealed these are the final measurable success in the management scope. 1.2 Customers Another problem as reported by Lu et al. (2005) is user acceptance. The customer of PMA is defined as the M&E users who will depend on the portable devices as their daily support. Hence, a new approach is generated to look from the user’s perspective on PMA’s capacity assessment. This study aimed to exploit the new, mobile empower user-group because the engineers are to deal with dynamic aircraft maintenance, to meet with the stakeholders’ expectations and measurable. Thus, gives rise to a new scope of data collection for process mapping. A group of M&E companies and airlines is formed to take on this challenge in order to explore from system-application combination, in the hope of creating a set of guided rules for PMA adoption. 1.3 A shared dream Both the stakeholders and the users, OEM PMA data suppliers and technology providers would share a common dream: This can be shown in Figure 1 as that in the future all information to support the maintenance operation will be readily accessible through a PMA as small as a button. We anticipate technology progression will eventually deliver a “dream” PMA working platform for us. The design of the electronics is beyond the scope of this research but the context of information and data management is. The question is on how should the stakeholder and the users to expect the PMA performance because it carries a lot of productivity sense for successful adoption of PMA at any stage of development with any shape of device. The hypothesised PMA in Figure 1 may seem distant but the concept of complete information availability for decision making in real time and in portable configuration

Systems and application development 787

JMTM 17,6 Engine 4 control fault

788 Health monitoring Maintenance tasks

Figure 1. Futuristic PMA for aircraft maintenance

Service providers

ERP, suppliers

Fault message Trouble shooting Remedial actions

has been taken root for a long time, as depicted by various researchers (Miller, 1990; Phaal et al., 2001; Davies, 2000; Nunn, 2001; Quah, 2002). Since, it is always critical for front line maintenance business units to work exactly to the airframer’s specifications even to bolts and nuts details – no room for mistake. Some ramp maintenance sections employ a van to carry all maintenance documents around the airport similar to a study by Zografos (2002). With existing mobile electronics, we want to construct a research scope to answer the two basic questions appropriate to any configuration of PMA to come: (1) If all information are available on the fingertips of the users, what linkages will PMA make a difference for maintenance processes? (2) If PMA concept is workable for all maintenance process, its effectiveness in this critical process must be demonstrated. The burden of proof falls to researcher who would regard the process as chain of process value, workstation value and at last the value for the whole production network involving all the parties. The system developers and the users should work together under the PMA research scope to satisfy the ultimate performance of the new process. True, today’s information and communication support technology is insufficient as per Lee (2003) to deal with the increasingly tasks demand with the higher data flow and varied digital formats in order to handle the intricate process steps and decision making (Chan et al., 2002) on the spot. Many approaches had been taken, however, there is no fixed rules. Hence, a special methodology from users perspective in critical production process is conceptualised for effective referencing, recording and data transmission on a seamless input/output operable platform. 2. Literature review System development projects usually commenced on a cost-effectiveness initiative as per Pahl (2003), thus to anchor on a revenue model. Once the systems spec are well defined that a user-centred design (Galer, 1996) is employed to ensure the application effectiveness. We start from Mennecke’s (2003) concept for development agents and technology providers’ understanding on specific application needs, with a support on highly specialised core competencies referring to Holloway’s (2003) “list of necessities”

for aircraft maintenance economics. The choice of developmental rules or particular paths refers to Phaal et al.’s (2001) technology roadmap with its effect on systems development and impact to the value generation configuration as per Afuah (2001). Then a search for the likelihood on systems-application insights from existing research gaps to create the model for data generation by user groups for this alternative concept as illustrated by Mennecke (2003). 2.1 PMA applications dimension 2.1.1 Core competencies, information and systems. A handful of studies (McBratney, 2003; Zografos, 2002; Fletcher, 2002; Ogaji, 2002; Rao, 2003; Greenough, 2001; Chan, 2002) illustrated the use of information technology and systems (IT/IS) in contract manufacturing, service and product development business, not only to monitor and measure operational effectiveness but to create greater common value. Chan (2002) reported that information portals and M&E performance measurements conceptualised towards data necessitated for business process and network agent, as depicted by Mennecke (2003), management information forms the basis to drive real time data capturing. Gunasekaran et al. (1997) depicted key issues of IT/IS supported production network players to access critical operational information with no time lag. In fact, with a long-term contract between two parties in a context for the “contract engineering and logistics services (CELS)” a high degree of collaboration should exist for co-development and co-defining the PMA systems for the service providers and the customer airlines. The question is who to initiate this non-existing but hand-shaking PMA system between the CELS companies? 2.1.2 Equipment perspectives – military vs airlines applications. Existing equipment and configuration affects research design as per Lu et al. (2005) to conceive PMA-enabled airline maintenance operation (Figure 2). A new breed of airline maintenance expertise and application advancement is enticed to improve the existing

Systems and application development 789

Process Performance

Remote workstation

User Adoption

Application Level

F-16 EDNA sys

M&E user

Wearable

PDA

Tablets

SPORT sys

F-22 Raptor sys

Figure 2. Equipment perspective PMA-process performance

JMTM 17,6

790

paper-based process. Availability study on military applications with 33 cases of PMA applications as per Bapst (2002), currently only two appeared in commercial airlines. The most reported application so far is the EDNA system for F-16 aircraft diagnostic, preventive and responsive maintenance support. Users quickly realised that commercial M&E business exhibited the following features that are different to military applications: . more dynamic business strategy, less on fleet standardisation, relatively faster equipment rollover rate; different service nature; . more on collaborative model between CELS players than the governance model in military for system adoption and substantiation; and . more on cost-effectiveness, user preference than the disciplinary and ordinance practices. A high extent of customisation is expected if the above equipment is to be adopted in any stage of mobile computing practices. The top level of adoption for improving trend of process performance, as conceptualised in the equipment/application development model (Figure 3). 2.1.3 Commercial airframer’s software and data support. Presently, Boeing and Airbus take ownership for publication of the digital maintenance documents. Airlines and M&E companies may adopt airframer’s software directly, but certain linkages required to facilitate interchange of key operational data between local users. The users and system developers are encouraged to explore the generic Boeing and Airbus data formats as follows: (1) AIRN@V. The Airbus software is conceived for maintenance support (Airbus, 2000) in all dimensions. A web-based or load-able system that provides:

Process Performance

Systems Appreciation

Controlled maintenance information & data M&E info support

Ge ne ric PM A da ta

Systems Level

Figure 3. Generic airframers’ PMA data concepts

Fleet level

Ge ne ric PM A da ta

Ge ne ric PMA da ta

Gene ric PMA da ta

Aircraft level

Engine level

Sub-assembly level

Piece part level

supports real time receipt and management of messages from aircraft on-board system through airborne data communication allows mechanics to prepare a maintenance action list; . OEM support through the AIRMAN connection; and . advanced search functions within the Airbus network. (2) Boeing PMA. The Boeing equivalent (2003) includes Boeing initial maintenance documents and later more advanced allowable configuration manager (ACM). This mobile support gear is designed around the self-sufficiency concept to reduce the time and effort needed to research, manage, source, order, and track parts thus shortening process time. Forming procurement lists for the parts with interchange-ability, engineering drawings, service bulletins, and maintenance procedures. .

The airframer’s solutions are good for specific aircraft types (but not include the competitors’ model). Developer agents need to integrate the software into M&E’s systems through separate sensoring, type-formatting, customising and updating. The developer’s (and the user’s) dilemma (similar to 7 , 9 cases) is how to translate the right portion of back office capacity to the front line so that the anticipated performance level is demonstrated, as top level for information/system perspective. 2.2 Mobile system development dimension 2.2.1 Nomadic network advancement. The methods for wireless development follows product sustainability and scalability O’Brien (2004), but we still observe the rules of system development when it is introduced to the aircraft maintenance industry: (1) System development and life cycle (SDLC) through SSADM in, 1970 and its robustness vs inherent mobile flexibility and programming. (2) Waterfall model to help developers to captures business process and system specification development (BPSSD) potential. (3) Spiral model and sustainable development model to ensure adaptive adoption of the gear-software and data support. (4) Rapid application development and rapid prototyping to test out the application data for the required systems. (5) Agile system design for the three levels of expectation. According to So (2004) any nomadic system development should consider the widest possible involvement of the users. Hence, as depicted by Chan (2002) rigid systems development theories from the listed methods (1)-(5) above mentioned, are to be selectively adopted but not necessary in chronological order. 2.2.2 Exploring the technology roadmap. In the course of technology adoption, nomadic computing development has reached maturity hence a conceptualised operation can designed as Phaal et al. (2001) advocates that apart from form the electronics itself the process-application and social-economical issues to be considered: . Market. PMA forms the basis of networked production hence it should not be limited to moving the known back office operations into the front line, a critical-processes study must be launched (Phaal et al., 2001).

Systems and application development 791

JMTM 17,6

.

.

792 .

Research and development. PMA systems can be expanded into the nomadic operation environment for different stage of breakthroughs, e.g. light-weightness. With a predictive trend in product availability, it is more important for potential users to clarify what to be benefited, instead of how to use the technology when it comes. Product and service. Forward thinking users and companies may wish to demonstrate PMA feasibility. An early success from initial test results may accumulate application intelligence through structured demos to the potential users. Sustainable development. The creation of M&E network influence.

The argument for how to sponsor a commercial user in anticipation of the hard/software couple through transparent and convenient form of wireless PMA support is presented. The developer has to conceptualise and design such an experiment to challenge or assist existing process. In that to explain why PMA is selected, amongst other things, how should the project be resourced and funded for high priority development. 2.2.3 Narrowing the gap between value process and value shop. Parolini’s (1999) proposed to group all economic activities for the creation of a value network in three layers: supporting, tools creating and transaction activities. Chan (2002) depicted that the production network is expected to achieve the targeted improvements in three groups of activities: technology adoption, IT implementation and performance activities. A greater value sum is envisaged and shared with other players to attain the intelligence in application. To achieve this, Afuah (2001) hinted that infrastructure builder should study the value of the process first, integrate its requirements and embed the activities in a demonstration of the mobile system prototype for a group of critical activities. So that strategic functionalities are illustrated as the form of a “value shop”. 2.3 Summary of literature review Porter’s (1999) competitive organisation theories has led M&E industry into finding opportunities between suppliers, CELS partners and competitors to re-evaluate the maintenance process and position. No one particular configuration has become domination other than simple business competition. This paper attempts to demonstrate a process with the above supportive literatures, whereby an alternative systems-application intelligence gathering could debate for the nomadic network adoption in this rich-content, time critical precision production arena. 3. Methodology The system-application approach adopted advanced use of the devices and systems in matching process with network. The suitable hardware-software couple would be provided either in partial prototype or narration. Since, no standard has been set for such nomadic applications (as oppose to a standalone lap-tops) data collected should be subject to correlation test (a priori analysis ref. Adamo (2001)) between the user preference and process performance parameters. Early adoption of Galer’s (1996) list of user-centred design is a key feature for this methodology, i.e. without ask the stakeholders to define and freeze every features of the PMA application (it probably could not due to commercialised mobile products’ rapid

updating almost on quarterly basis). Mennecke’s (2003) performance traits were then used as digital agent and process agent. Selected network users will be invited to test the nomadic computing device and for comment. Work process correlation to soft/hardware would feedback to design of the device, reshape it based on user appreciation levels. Finally, overall effectiveness towards business process, spec refinement, software programme refinement and de-bugging. As oppose to the classic IT performance dimensions (O’Brien, 2004) one could break down the huge block of back office operations into smaller chunks, and then conceptually reform them together to develop a model as shown in Figure 4. The prototype definition is reformed and clarified. It aims to: . successfully pass the user go/no-go gauge; and . establish the three-layers (and sub-layers) for correlation analysis. 3.1 First layer – L1 generic maintenance value process Understanding the CELS business relationship helps resolving system level information format bias. Layer 1 gathers multiple tier suppliers, airframers and CELS information provision to result in a format biased (or ownership defined) PMA data. Certain critical CELS information (Greenough, 2001) is originated from the firstand second-tier players (original equipment manufacturers OEM’s). Key protocols for the existing digitised documents and operational data (Figure 5) and responsibilities (Table I) for respective CELS players are under scope of layer 1. In reality, global production network drives each player to work on no less than five types of “islander systems”: enterprise resource planning (ERP) data, in-house documentation, performance indices and work scheduling, etc. Process agents define how and where information from different players is to be positioned and prioritised in the limited devices. 3.2 Second layer – L2 nomadic support in critical process The network IS formats are then subject to layer 2 (design) which stress on real-time, compressed, accurate and communicate information between off-sites users as depicted by Fletcher (2002). So that inter-firm linkages bring together various trades effortlessly. The development agent integrates and rationalises nomadic specifications supporting specific applications, in airline maintenance service case, the flight despatch critical software support. This layer is equipment-centred and is catered for different and complex operational situations. A prototypic equipment is designed for data collection (named Crystal1) and distributed for the ramp engineers to trial use without a lot of training other than showing what contents and linkage are built-in. 3.3 Third layer – L3 user appreciation The last layer is the capturing of the users feedback for the layers of L1 and L2. This is designed to analysis L1 and 2 layers data, in which the user experience is rated. In order to narrow the scope, three questions have been set for those who will trial use the Crystal1 test rig: . Is the business tie into the mobile device that you are closely using or most familiar with? . Is a direct process improvement encountered in regular tasks? . Is training an important part of the PMA device for the critical process?

Systems and application development 793

Figure 4. Systems-application intelligence layers ss Time eline y Tim rrenc ncy Fo rm Cu eq ue rity Cla Fr riod Co tailnte Pe DeAcc eruranton Fo rm Ord taticy Rel seneva rity Pre Cla Com ncy ail dia Det Co nciple ten Me ess Sco sen ess Order tation pe sen Per Pre for ma nce Me dia

Fo rm rity Cla De tail Order tation sen Pre dia Me

L3 User appreciation

Fo rm rity Cla ail Det Order tati on sen Pre dia Me

ess Time lin y Time rrenc ncy Cu eq ue Fr riod Pe

mp ess Time lin y Conci letenes Time rrenc ncySco seness s Cu eq ue Per pe Fr riod forma Conte Pe nce Accur nt Fo rm Rel acy rity eva Cla Comp ncy Conci letenes Detail er s Sco seness Ord tati on pe sen Per Pre forma nce Media

Rel acy eva ncy Co

Conte Accur nt

L2 Nomadicity

Ac cur Relev acy Comp ancy Co nci letenes Sco sen ess s pe Perfor ma nce

Cont Accurent Relev acy Comp ancy Co nci letene ss Sco sen ess pe Per for mance

e ess Tim lin y Time rre nc nc y e Cu eq ue ess Tim lin y Fr riod Pe Time rrenc ncy Cu eq ue Fo rm Fr riod Co rity Pe Cla n ten Ac t Detailcur acy Fo rm Re lev n der rity OrCo atio Cla l s ent anc PreCo mplet y De tai dianci en ess n MeSc sen Orderntatio ess op e e Peess Prese Ti m lin rfoy dia y Time rre nc rm anc Me Cu eq ue nc e d Fr rio Conte Pe nt

L1 Production network IS formats

Conte Acc nt ura Rel cy eva Com ncy Co nciple ten ess Sco sen ess pe Per for ma nce

ess Time lin y Time rre nc cy Cu eq uen Fr riod Pe

Systems-application intelligence transmutation

PMA tech demonstrator

Prototype design

PMA spec analysis & selection

Applications Go or No-go gauge

carefully tailored on-board hardware/software couple

seam less nomadic operation with suitable scale of equipment

optimised CELS connectivity between front line & boo interface

PMA spec reform for critical process testing

794

e ss Tim eline cy m n i T urre ncy C eque Con Fr riod od tent Pe Acc m urac For Rel y rity e Cla l Com vancy ai p t l e e Con tene D ss cise er n d nes Sco Or entatio s p es Per e r P form ia anc Med e

IT performance facets (O’Brien [14])

JMTM 17,6

Maintenance Service Provider

Spare and part repair

Component Suppliers

2nd Tier

& Tech Support

Build-up, modify

Maintenance & Repair

Airframe Manufacturer

Spare supply & part repair

1st Tier

Serve

Fly &

Passengers & freights

a circle denotes the relative closeness of information support format & systems definition towards the party forming the contract service agreement.

Airline operators

arrow denotes service direction & material & information flow which is two way through contractual agreement. This network two stages: a) new production & b) in-service of the aircraft.

Systems and application development 795

Figure 5. CELS task network and information configuration

JMTM 17,6

796

CELS player

Responsibility in aircraft maintenance Bias of IS format

Airline operators

Charter the use of aircraft fleets and assets such as spares/support equipment and ensure airworthiness Build the aircraft and provide after-sales maintenance technical support and modification Provide new spare parts for replacement and repair/maintain used parts for exchanging Maintain aircraft to flight schedule and work to safety and reliability standards and international regulations

Airframe manufacturer Component suppliers

Table I. CELS player responsibility vs IS format

Maintenance service provider

CELS context management by contract Connectivity to the original spec (airframers) Commonality to various specs (OEM’s own) Contractual requirement, usability and generality

The Crystal1 device is used, either directly or working in-lieu of the engineer’s focus to capture the maintenance data or information needed for the test. Collected data are to be tested by a priori method and fed back to CELS network companies for early success demonstration, as well as to refine Crystal1 design and features incorporation.

4. Case study – PMA supported on-time performance This case study captures selected critical network process (airline flight despatch) using the three layers of systems-application intelligence embracing various equipment and software couples to support fast response and sensitive maintenance tasks accomplishment. 4.1 Background An international airline which operates a fleet of 50 jumbo jets adopted the CELS network approach to manage outsourcing tasks in different companies with a view to create: . supply chain logistics; and . technical competence management. The company is keen to operate in good standard of on-time-departure, hence Crystal1 is used to: .

.

.

validate application of mobile technology in the CELS network with focus on ad hoc maintenance and situation management for L1 design; translate lessons learned in resolving user interface and design issues as application intelligence so that the result would allow the designers to explore L2 refinements; and capture general features of mobile application by the users in real life simulation and subject L3 analysis.

4.2 PMA enabled “operating theatre” concept process management Selected in-house and outsource functional users gathered to evaluate the given mobile gear. Flight despatch data are collected from the Hong Kong’s CLK Airport’s maintenance activities across CELS network for transit maintenance. The concept of “maintenance theatre” is initiated to enable PMA system as an “actor” in different critical situations. A typical dynamic scenario involving is depicted as follows: . Arrival. A B747 freighter was scheduled to arrive at bay carried 15 deferred tasks. Already 20 minutes late on bay, the aircraft would arrive for unloading. (L1: connectivity between CELS companies communication for the right time to start of work). . Inspection. Engineer revealed that the condition of a certain system was marginally acceptable but needed authorised relief. An oral report with picture and description would be sent to the authorised party (airline). Relevant approved document accepting the proposed action is expected (L2 initiation). . Inbound. Also, defect abc is registered during flight. Before the aircraft was landed some information from the aircraft’s central maintenance computer (CMC) was downloaded to the ERP system for rectification analysis (L2 PMA efficiency measurement). . Trouble shooting. The avionics technicians trouble shot the problem in cockpit and request the component change. After the defective computer was changed, the system is tested through a PMA (L2, nomadic computing efficiency measurement). . Loaded. Aircraft door closed for departure, a fault xyz message appeared. An inconvenient time for defect to crop up, the engineers on the spot needs to know the root cause and any misleading faults would be time wasted (L3 details). . Decision. A delay is anticipated all parties in airline and service providers meet on the mobile device for consent (as oppose to the current practice of discussing the details over walkie-talkie radio), and to acquire a resolution. Remedial action might take several hours, management of the logistics activities became tense (L3 comparison of voice to data communication). PMA data is transferred to headquarter in real-time, expertise support enhance the trouble-shooting. The rectification task is carried out in minutes only. . Data record. Data capturing, as opposed to the existing method of tape recording the voice communication, would be of interests here. The data for fault identification and logistics activities would be captured for every bit of data transaction (L3 overall PMA effective comments). The team initially used a simple PDA to test the above process. With sufficient data captured the PMA equipment is enhanced such that a device created and refined, the Crystal1 (Plate 1). Even though it was not yet close to the “dream” device in Figure 1 but significantly improved from existing devices as shown in Figure 2. This prototype validated the critical process as a demonstration for the CELS network until the hard and software are sufficiently matched for the critical process support. Key design features were articulated to perform as the reference equipment as it evolved:

Systems and application development 797

JMTM 17,6

798 Plate 1. PMA prototype – Crystal1 illustration

4.3 CELS players process changes driven by PMA The purpose of the experiment was to show case the concept of PMA enabled maintenance theatre activities would work and be visualised under the Layers 1 to 3 data encapsulation scope. Usability in the selected CELS revealed no difficulties. The technology roadmap was trusted at this development stage (in-lieu of the developer and researcher along with the engineer), thus could be moved to the in-service stage (go-solo, just for those trained engineers who works independently). A user focus group was assembled to test the functionality and evaluate the improvement in operational efficiency resulting from the use of PMA systems. The net effect for L1 initiated PMA changes is considered as paradigm change. The reduced overall costs and improved content accuracy and consistency were reported. Of note, detail workflow analysis is subject to a separate research project focused on operational research methods for process improvement. 5. Discussion Having completed the Crystal1 prototype using a three layers approach (Figure 6), L1 generic value process, L2 nomadic application and L3 users appreciation, sufficient data is collected for analysis on the effectiveness of this approach in capturing the intelligence need to design something workable in the CELS process. Even though nomadic is the environment enabler (or had been enabled) the first and third layers remain crucial to the early success. The knowledge collected for workstation constraints helps to overcome difficulties in updating work contents between remote sites and back office operation, of note, one remote site test for Crystal1 test extended to Osaka Kansai Airport with no difficulties reported. With Crytal1 solution as the datum, structured mobile XML/SGML information had been used to measure the following: . improved overall linkage of content management to core product/service measurement; . streamlined costs associated with content creation, management and delivery measurements; and . increased organisational agility and effectiveness by automating delivery measurements.

Boeing ACM in used L1b AIRMAN in used L1c CELS networked L1a Business unit strategy L3a

START: Contract

M&E services

Sensors tested L2a

Relative strong correlation paths

Maintenance site tested L2b

Systems and application development 799

Management buy-in L3b

Training instigated L3b

Other nomadic applications L2c

These settings tested PMA/software support for flight despatch. More data from the overall maintenance review would be needed. This is not only for the despatch phase but also for all working phases and maximum extension to OEM PMA data sources in web or load-able configurations. 5.1 Theatre management collaboration model The theatre approach to bring out cross-disciplinary issues to one point for conflict resolution promoted understanding of the “big picture” of operation. The importance of decision support for step Decision) in the Case stood out. Through this exercise, PMA captured extensive intelligence from business partnerships with complementary CELS organisations willing to test the nomadic network as well as to provide constructive feedback. The L2 study allows researcher to study technical conformity across large network areas and made use of secured wireless technologies including wireless LANs. On layer 3, it promoted effective use of resources via PMA in both the airlines and M&E firms because the purpose of systems-application for PMA adoption depended on a number of factors that are interwoven between the sub-layers. If strong correlation is identified, that helps both development and process agents as per Mennecke (2003) to strive for early success in the application. An a priori analysis was carried to find out the correlations between the elements defined in the three layers of systems-application intelligence and identified critical paths for early success factors in theatre management, as shown in Figure 7. The strongest correlation should initiate the start of process redesign for flight despatch improvement: 5.2 PMA vision created for M&E stakeholders With the successful demonstration, stakeholders envisaged that rich content PMA incorporate configuration control to manage and share customised data for specific fleet to attain significant efficiency improvement. Multiple formats linking to supplement the electronic flight bag (EFB) concept in the cockpit. Future issues can be assessed based on the prototype Crystal1:

Figure 6. A priori association analysis of three layers

Figure 7. Table of a priori association analysis for three layers 2%

– 5%

3 4% – 12%

– 4.5%

L1c.L2 a …

L2b.L3a

… L3a.L2c

–

56%

34%

73%

L1b.L2a

– –

L2 Nomadic

15%

–

L1 CELS

… L1a.L2a

SubLayers

Apriori Association Analysis

7%

standard

Weakest link

standard Strongest bon strong

Correlation comments

Airframe Manufacturer

Maintenance Service Provider

Component Suppliers

Early success process starts here

Airline operators

800

–

1%

– –

21%

44%

–

L3 User

JMTM 17,6

.

.

.

Matching more milestone points to business-system integration and demos in the technology roadmap (in practical terms, bigger screen, approximately 100 times more enriched information to be displayed). More precise sensoring, network, tasks in theatre view to confined details such as RFID/GPS for surgical precision logistics and technical control for all maintenance work (future linkage to airframers’ and ailrines’ own network-based and global-activation systems). Crystal2 development to explore the nomadic environment based on newer prototype to strengthen the mobile information display, and to connect with the value network (researchers’ next challenge).

Systems and application development 801

Sufficient data is provided for Value Process-to-Shop level of system-application development (Figure 8). Areas of improvement are defined that will require business process enhancement to align with mobile PMA adoption. 5.3 Limitations of systems-application method As a whole, this PMA systems-application test was carried on airline CELS players and a confined environment for the process of flight despatch merely. This test had not included down stream component suppliers or airframers’ of the CELS value network. Inclusion of these parties should be considered as the whole production network, hence the “value net”. In future studies more data should be captured in further dispersed OEM and outstations. Hence, a fourth layer of tier 2-4 suppliers can be explored to complete the whole CELS value network experimentation as depicted as follows 5.4 System-application performance improvement next phase . Analysed data supports that a traditionally efficient back office platform may not the ideal template to be copied for front line use. . The need for user preference to enhance the accessing speed in the mobile or nomadic computing environment would be different to that of the desktop interface.

Value Network Crystal 2 to demonstrate wider value net concept

Value Shop

Crystal1 to test the Sys-Appl'n concept

Value Process

Boeing & Airbus, other OEM's and extensive study through other M&E business and airlines. Critical linkage to other firms in a confined scope and processes Test of the PMA technology in the high value task such as critical process of flight departure.

Figure 8. Value configuration progression path

JMTM 17,6

.

.

802

.

Some users suggested the use of graphic interfaces as much as possible, which is becoming the key feature in Crsytal2 prototype development. The system-application research has started a new way of putting user preference data early into the adoption study, which generates a lot of queries into comparison of the research into possible electronic configuration, rather than the method of using it. The user defined application system approach is considered to be user-friendly, despite the confined scope of application.

The research results should be subject to extensive analysis for external parties (outside Crysatl1 user scope) in order to collect data from: . Boeing and Airbus, multiple tier suppliers who would comment on how researchers would use the OEM data by injecting with more M&E operational data, such as planning and logistics parameters. . Other M&E industry and service agents had not been included in the test should be invited. . Other airline operators who would consider PMA as an important system technology enabler for their future development. 6. Conclusion This system-application approach has successfully demonstrated two basic levels of value configuration as suggested by Afuah (2001). That is, the first level value chain for the maintenance tasks accomplishment. The second level the value shop in which the remote workstation for aircraft despatch is managed by the engineer-in-charge through advanced wireless support concept so that the PMA configuration will achieve the usability performance margin. Whereas the third level of value transaction (data, information, knowledge and material) must occur in the extensive value network, as defined by Afuah that Airbus and Boeing should be joined to provide the necessary data to link up the value network between the suppliers, M&E suppliers and the airline operators. With the escalated aircraft asset utilisation, streamline the logistics concerning repair, overhaul and modification to global scale has come a strategic importance. Stakeholders, managers and the users/customers have to attain performance targets and demonstrate unwavering task accomplishment standards in dynamic real world scenarios. This PMA study studied a new paradigm to exploit the possibility how high resolution PMA maintenance data to support critical process, especial from process and system performance perspectives. Data is collected in a mobile-enabled workforce empowered environment in search of the mobile device to support: . optimised production network process; . seamless nomadic operation for the workshop/workstation with limited inter-firm level; and . tailored made hard/software to entice the user-centred design. We have generated a set of guided rules for PMA development especially in the complex network-nomadic-appreciation intelligence gathering for early success.

It may take some time to put the ideal PMA platform (Figure 1) into practical service. Under this confined scope of mobile solutions research we instigated better linking the “islander systems” so that they could connect processes and to link up existing specialisation. It is conceivable that other dimensions of service such as aircraft interface, passenger or airport support can be instigated for further experimentation.

References Adamo, J.M. (2001), Data Mining for Association Rules and Sequential Patterns: Sequential and Parallel Algorithms, Springer, New York, NY. Afuah, A. et al. (2001), Internet Business Models and Strategies, McGraw-Hill, Boston, MA. Airbus Industries (2000), “Airman introduction”, Airbus Journal, Vol. 2000 No. 11, pp. 27-36. Bapst, G. et al. (2002), Portable Maintenance Aid, LG005T2, Logistics Management Institute, July. Boeing commercial airplanes (2003), “Portable maintenance aid: maintenance on your finger tips”, available at: www.boeing.com/commercial/ams/digital/pma_pmaskd.html (accessed February 2004). Chan, M. et al., (2002), “Information portal and the impact on e-business”, International Journal of Production Economics, Vol. 15, pp. 234-67. Chan, M. et al. (2005), “Development of performance management system for aircraft maintenance”, working paper, International Journal of Production Economics, November. Davies, A. and Brady, T. (2000), “Organisational capabilities and learning in complex product systems: towards repeatable solutions”, Research Policy, Vol. 29 Nos 7/8, pp. 931-53. Fletcher, J.D. et al., (2002), “Effectiveness and cost benefits of computer-based decision aids for equipment maintenance”, Computers in Human Behavior, Vol. 18 No. 6, pp. 717-28. Galer, M. et al. (1996), Methods and Tools in User-Centred Design for Information Technology, North Holland, Amsterdam. Greenough, R. (2001), “Development of a digital manual for a manufacturing system – a case study”, Integrated Manufacturing Systems, Vol. 12 Nos 6/7, pp. 387ff. Gunasekaran, A. and Nath, B. (1997), “The role of information technology in business process reengineering”, International Journal of Production Economics, Vol. 50 Nos 2/3, pp. 91-104. Holloway, S. (2003), Straight and Level: Practical Airline Economics, 2nd ed., Burlington, Ashgate, Hampshire. Lee, J. (2003), “E-manufacturing—fundamental, tools, and transformation”, Robotics & Computer-Integrated Manufacturing, Vol. 19 No. 6, pp. 501-7. Lu, J. et al., (2005), “Personal innovativeness, social influences and adoption of wireless internet services via mobile technology”, The Journal of Strategic Information Systems, August. McBratney, A.B., Mendonc¸a Santos, M.L. and Minasny, B. (2003), “On digital soil mapping”, Geoderma, Vol. 117 Nos 1/2, pp. 3-52. Mennecke, B.E. (2003), Mobile Commerce Technology: Theory and Application, Idea Group Publication, Hershey, PA. Miller, D., Mellichamp, J. and Wang, J. (1990), “An image enhanced: knowledge based expert system for maintenance trouble shooting”, Computers in Industry, Vol. 15 No. 3, pp. 187-202.

Systems and application development 803

JMTM 17,6

804

Nunn, S. (2001), “Police technology in cities: changes and challenges”, Technology in Society, Vol. 23 No. 1, pp. 11-27. O’Brien, J.A. (2004), Management information systems: managing information technology in the business enterprise, 6th ed., McGraw-Hill, Boston. Ogaji, S. (2002), “Novel approach for improving power-plant availability using advanced engine diagnostics”, Applied Energy, Vol. 72 No. 1, pp. 389-407. Pahl, C. (2003), “Managing evolution and change in web-based teaching and learning environments”, Computers & Education, Vol. 40 No. 2, pp. 99-114. Parolini, C. (1999), The Value Net: A Tool for Competitive Strategy, Wiley, Boston, MA. Phaal, R., Farrukh, C. and Probert, D. (2001), “Technology roadmaping: linking technology resources to business objectives”, available at: www-mmd.eng.cam.ac.uk/ctm (accessed February 2001). Porter, M. (1999), On Competition, Harvard Business Press, Cambridge, MA. Quah, J.T-S. and Lim, G.L. (2002), “Push selling – multicast messages to wireless devices based on the publish/subscribe model”, Electronic Commerce Research and Applications, Vol. 1 Nos 3/4, pp. 235-46. Rao, J.S. (2003), “Condition monitoring of power plants through the internet”, Integrated Manufacturing Systems, Vol. 14 No. 6, pp. 508ff. So, H. and Chung, W. (2004), “Mobile IT infrastructure in value network development: a case study of property management business”, International Journal of Production Economics, Vol. 2, pp. 2-15. Zografos, K.G., Androutsopoulos, K.N. and Vasilakis, G.M. (2002), “A real-time decision support system for roadway network incident response logistics”, Transportation Research Part C: Emerging Technologies, Vol. 10 No. 1, pp. 1-18.

Further reading Aerts, A.T.M., Szirbik, N.B. and Goossenaerts, J.B.M. (2002), “A flexible, agent-based ICT architecture for virtual enterprises”, Computers in Industry, Vol. 49 No. 3, pp. 311-27. Bellotti, F., Berta, R., De Gloria, A. and Margarone, M. (2003), “MADE: developing edutainment applications on mobile computers”, Computers & Graphics, Vol. 27 No. 4, pp. 617-34. Fitzgerald, G. (2003), “Research challenges in information systems”, International Journal of Information Management, Vol. 23 No. 4, pp. 337-44. Giannopoulos, G. (2004), “The application of information and communication technologies in transport”, European Journal of Operational Research, Vol. 152 No. 2, pp. 302-20. Kiritsis, D., Bufardi, A. and Xirouchakis, P. (2003), “Research issues on product lifecycle management and information tracking using smart embedded systems”, Advanced Engineering Informatics, Vol. 17 Nos 3/4, pp. 189-202. Klusch, M. (2003), “Information agent technology for the internet: a survey”, Data & Knowledge Engineering, Vol. 36 No. 3, pp. 337-72. Majstorovi, V. (1990), “Expert systems for diagnosis and maintenance: the state-of-the-art”, Computers in Industry, Vol. 15 Nos 1/2, pp. 43-68. Mora, P., Baldi, P., Casula, G., Fabris, M., Ghirotti, M., Mazzini, E. and Pesci, A. (2003), “Global positioning systems and digital photogrammetry for the monitoring of mass movements: application to the Ca’ di Malta landslide (Northern Apennines Italy)”, Engineering Geology, Vol. 68 Nos 1/2, pp. 103-21.

Sein, M. et al. (2001), Contemporary Trends in System Development, Kluwer Academic/Plenum Publishers, New York, NY. Sherif, Y. (1982), “Reliability analysis: optimal inspection and maintenance schedules of failing systems”, Microelectronics and Reliability, Vol. 22 No. 1, pp. 59-115. Swartz, J. (2003), “Security systems for a mobile world”, Technology in Society, Vol. 25 No. 1, pp. 5-25. Veryard, R. (1992), Information Modelling, Prentice-Hall, Englewood Cliffs, NJ. Corresponding author Walter W.C. Chung can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

Systems and application development 805

The current issue and full text archive of this journal is available at www.emeraldinsight.com/1741-038X.htm

JMTM 17,6

806

E-business capabilities model Validation and comparison between adopter and non-adopter of e-business companies in UK Khalid Hafeez and Kay Hooi Keoy Bradford School of Management, Bradford University, Bradford, UK, and

Robert Hanneman Department of Sociology in the College of Humanities, Arts and Social Sciences, University of California, Riverside, California, USA Abstract Purpose – The purpose of this paper is to present a conceptual framework to evaluate e-business strategic capabilities using structural equation modelling (SEM) approach. Design/methodology/approach – The paper identifies three e-business capabilities, namely business strategy, supply chain strategy and e-business readiness. These capabilities are further decomposed under technology, organization and people dimensions to assess their contribution for business effectiveness. A questionnaire is designed and implemented using SEM technique. Survey data from 143 firms from the UK are collected to test our theoretical model. In particular, we have tested a positive, mediating/reciprocal relationships among multidimensional measures of business strategy, supply chain strategy and e-business adoption. Further hypotheses are developed to evaluate a direct positive impact of e-business on company’s performance. Findings – This empirical analyses demonstrate several key findings: success of e-business in UK firms is attributed to the strong positive co-relationship of supply chain strategy to business strategy and to e-business adoption; within the technology-organization-people dimensions, e-business adoption and business strategy emerges as the strongest factors for the company’s performances for the adopter of e-business group, whereas supply chain capabilities and business strategies is relatively a stronger contributory factor towards business success for non-adopter of e-business. Originality/value – It is expected that the results from this study will provide useful guidelines for companies to assess their strengths and weaknesses towards adopting an effective e-business implementation strategy. .

Keywords Corporate strategy, Supply chain management, Electronic commerce, Modelling, United Kingdom Paper type Research paper

Journal of Manufacturing Technology Management Vol. 17 No. 6, 2006 pp. 806-828 q Emerald Group Publishing Limited 1741-038X DOI 10.1108/17410380610678819

Introduction It is understood now that e-business is more than just another way to sustain or improve existing business practices. Where a few researchers suggest that e-business is a “disruptive” innovation that is radically changing the traditional way of doing business (Lee and Kalakota, 2001), on the contrary many other highlight the evolutionary aspects of e-business adoption (Ross et al., 2001). There are examples where companies have used business tools such as total quality management (TQM) and business process reengineering (BPR) while introducing e-business in the organisation. These tools allow companies to automate processes, integrate systems, and work towards customer intimacy. As Ross et al. (2001) suggest that migrating to

e-business involves exploiting existing processes through information rich channels, expanding core processes to include adjacent businesses, and extracting management attention from these processes. A recent survey suggest that two third of the UK businesses are online, while further growth is forecasted (UK Online, 2002). There are indications that the larger companies adopt twice as many e-commerce activities compared with small and medium-sized enterprises (SMEs) (Haig, 2002; Simpson and Docherty, 2004). In addition, the UK government has recently acknowledged that there is a slow take-up of e-commerce amongst SMEs, particularly amongst micro-businesses (UK Online, 2002). The main reason cited for this slow adoption is ignorance about e-commerce benefits and a shortage of relevant skills (DTI, 2002). The work reported in this paper was set out to develop an in depth understanding how UK companies have prepared for e-business adoption. A number of key factors contributing towards e-business success are identified. A comparative hypothesis testing methodology is proposed to investigate the problems and barriers in implementing e-business strategies across multiple organizations. This research also aims to develop a measurement tool that academic and practitioner can use to identify existing gaps between the firm’s current business practice and its desired e-business implementation strategy. The paper is organised as follows: some relevant literature is reviewed in first section. Conceptual e-business capabilities framework with underlying technology-organisation-people (TOP) dimensions is discussed in the following section. Research method using multiple group comparison between adopter and non-adopter of e-business groups and the outcome of the structured equation modelling (SEM) analysis are given, respectively. Lastly, discussions on the theoretical and practical implications of these results, as well as limitations of this study are given. Literature review A few authors view e-business as one of the most controversial research area (Zhu et al., 2003, 2004). Despite the burst of the dot-com bubble few years ago, many companies are still continuing to deploy e-business extensively in their business operations. There is overwhelming evidence that firms such as Dell, Wal-Mart and Capital One have achieved improvements in their operational efficiency and supply chain partners intimacy by integrating e-business into their business models. Many large and SME companies have developed internet-enabled initiatives to strengthen online connection with customers, disseminate product information, facilitate transactions, improve customer services, and manage inventory via electronic links with suppliers. The dominant strategy among retailers is to achieve direct access to customers via interactive web technologies. There are reports indicating that due to the fear of lagging behind the technology curve, many firms have anxiously engaged in e-business initiatives without deriving any benefits (Martinsons and Martinsons, 2002; Barua and Mukhopadhyay, 2000). There is evidence that embracing technology can result in competitive advantage in the marketplace (Levitt, 1983). Internet technologies offer opportunities for instant international market access, as well as improve domestic market performance for companies (Keogh et al., 1998; Caviello and McAuley, 1999; Ian et al., 2004). By using internet, customer relationships can be re-engineered resulting in a more cost efficient one-to-one relationship (Webb and Sayer, 1998). Therefore, there is a view that

E-business capabilities model 807

JMTM 17,6

808

widespread business adoption of the internet is needed in order to reach critical mass for internet commerce (Poon and Swatman, 1999; Ian et al., 2004). Also there are guidelines that it is the longer-term benefits that should drive e-business development, rather than any short-term gains. (Ian et al., 2004). In the UK industrial climate, SMEs are increasingly making use of e-commerce with almost 92 percent of medium seized firms and 62 percent of small firms connected to internet access (Oftel, 2002). However, the SME sector is traditionally characterized by high failure rates, where failure rate noted to be six times higher for small companies compared to large counterparts (Storey, 1994). With the significant impact that e-commerce can have, such failure rate may increase if the UK SMEs do not develop efficient e-business implementation (Daniel, 2003). Therefore, many authors have suggested that the integration of internet with existing system is to be treated as an essential factor for e-commence effectiveness (Keeling et al., 2000; Melymuka, 2000; Haapaniemi et al., 2000; Von Hoffman, 2001). At present, much of the existing e-commerce literature relies exclusively on case studies, anecdotes and conceptual frameworks. Only a few studies use empirical data to characterize the internet-based initiatives, or gauge the scale of their impact on firm performance (Zhu et al., 2003, 2004; Brynjolfsson and Kahin, 2000). There is also a lack of theory to guide the empirical work (Wheeler, 2002). Many authors argue that the literature is weak in making the linkage between theory and measures apart from subjecting proposed measures for empirical validation for reliability and validity (Straub, 1989). E-business capability framework In this section, we proposed salient features of our conceptual framework. First we discuss TOP model in the context of e-business. Then we develop an appropriate e-business capability (EBC) model by discussing business strategy, supply chain strategy and e-business adoption flexibility for a company. Subsequently six hypotheses are proposed to be tested to show the relationships of TOP model with business performance employing structural equation modelling (SEM) technique. Results of the survey analyses are then discussed. Technology – organization – people (TOP) dimensions Many studies suggests that even in a traditional supply chain environment, many companies have yet not realised the full technological integration of the available office technologies and software tools such as material resource planning (MRP), distribution resource planning (DRP), and enterprise resource planning (ERP), etc. Steven as early as 1989 had advocated that in order to achieve full integration (from a baseline to external integration as illustrated in Table I); companies need to focus on technological, organizational, and people dimensions within and outside a company. Today many firms are still at the early stages to fully utilize the business opportunity and improvements offered by the internet technology. Therefore, we would argue that Steven’s (1989) supply chain integration framework is still fully applicable where companies want to move from a traditional business to e-business. We would argue that the three identified dimensions, namely; technology, organisation and people (TOP) are also well suited for studying e-business evaluation and implementation. Das et al. (1991) commented that technology has emerged as a flexible and versatile information attainment and processing capability which is essential to reduce

Stage of supply chain integration

Supply chain characteristics

Technology integration

Baseline

Organisation integration

Functional integration

Reactive short-term planning. Fire fight. Large pools of inventory. Vulnerability to market changes Emphasis still on cost, not performance. Focus inward and on goods. Reactive towards customer. Some internal trade-offs All work processes integrated. Planning reaches from customers back to supplier. EDI wide used. Still reacting to customer Integration of all suppliers. Focus on customer. Synchronized material flow. SC covers extended enterprise

People integration

Internal integration

External integration

Source: Stevens (1989)

the response time required by a company. In the context of e-business application the technology provide the essential capabilities and process applications. This may be reflected in terms of value management with regards to processing; provision of electronic exchange capabilities, and management of database (Croteau et al., 2001). There are studies showing that many organizations have pursued external integration leading to detrimental consequences by ignoring the organization dimension (Barratt and Green, 2001; Fawcett and Magnan, 2001).Organizational dimension can be defined as the choice pertaining to a particular configuration and internal arrangement intended to support the organization’s chosen position in the market (Morton, 1991). According to Stevens (1989), organisation flexibility is necessary to move towards internal integration of disparate operation functions. Also this enables an organisation to move towards an integrated MRP, DRP or a fully integrated ERP system. There are many studies suggesting an over emphasis on the technology, while the people issues have been completely ignore (Sabath and Fontanella, 2002; Barratt, 2002; Ireland and Bruce, 2000). Along with technology and organisational issues, senior management commitment towards e-business strategy and underlying performance measures are regarded having a strong impact on e-business success (Ontario, 2001). Kaplan and Norton (2004) describe people as organisation capital and company’s culture as its leadership, how aligned its people are with its strategy goals and knowledge sharing abilities of its employees. Achieving internal integration is not sufficient and could lead to create larger organization silos (Barratt and Green, 2001). According to Stevens (1989), attitudinal changes are necessary for a company to integrate with its customers and suppliers. For us, a fully integrated ERP solution, such as CRM, in only possible if companies’ mindset (or people) is willing to adapt to new organization and technological changes. Earl (2000) provides four e-business evaluation stages where organization, technology and people issues can be addressed appropriately. By utilizing Steven’s (1989) and Earl’s (2000) model, we believe that

E-business capabilities model 809

Table I. Stevens’ integration model

JMTM 17,6

810

these three dimensions make up the essential fabric for our e-business capabilities framework to study e-business value (Figure 1). E-business capability model As mention earlier, EBC model would include business strategy, e-business adoption and supply chain management (SEM) elements as shown in Figure 1. We define e-business adoption as the “readiness” of the organisation by having appropriate attitudes, skills, knowledge and technology to facilitate e-business operations. Again we would emphasize e-business adoption through technology, organization and people dimensions. As shown in Figure 1, all these components are inter-related, i.e. changes to one of these components will have ramifications on others. Porter (1985) developed a value chain model that highlights interdependent activities in the business where competitive strategies can be best applied and where information systems are most likely to have strategic impact. As information technologies developed, novel ways of business process redesign emerged. However, Porter (2001) argued that the key question is not whether to deploy e-business to take advantage of the internet technology, but how to deploy it. Gaining competitive advantage requires building on the proven principles of effective strategy either by operational effectiveness or strategic positioning. As Porter (2001) quoted: . . . the next stage of the internet’s evolution will involve a shift in thinking from e-business to business, from e-strategy to strategy. Only by integrating the internet into overall strategy will this powerful new technology become an equally powerful force for competitive advantage.

Technology Dimension

Organization Dimension

People Dimension

Business Strategy Business strategy ensures that organization, people and technological dimensions are a ligned to create a successful business. Company’s Business Performance E-Business Adoption The readiness of a company to introduce e-business processes taking into consideration or ganization, people and technological dimensions.

Figure 1. Proposed e-business capability framework

Supply Chain Management Supply chain strategies are aligned with the business strategy and e-business adoption taking into consideration organization, people and technological dimensions.

E-Commerce enhance company’s ability to create value position, increase revenue and performance improvement

SCM is defined as activities to be involved in the flows of materials (physical product flows from suppliers to customers through the chain and reverse flows via product returns, servicing, recycling and disposal), information (order transmission and delivery status) and finance (credit terms, payment schedules, and consignments and title ownership arrangements) in a network consisting of customers, supplier, manufacturers and distributors (Stevens, 1989; Hau, 2000). Clearly these flows cut across multiple functions and areas both within and outside a company; therefore, the internet and SCM are inextricably linked. The internet is a key enabler, providing radical improvement to the performance of many supply chain activities. Therefore, in our view an effective supply chain strategy must be part of overall business strategy (Keoy et al., 2002).

E-business capabilities model 811

Hypotheses formulation Many authors agree that, SCM should be given a higher level of strategic importance at the board room level (Maede, 1998; Philip and Pedersen; 1997; Damien, 2005). In addition, research has frequently identified that strategy development and business performance are to be inextricably link to reap the actual value proposition from the e-businesses (Rosenzweig et al., 2003; Vickery et al., 2003; Damien, 2005). Under the impression of our conceptual framework (Figure 1) and aforementioned discussions we postulated the following six hypotheses to test the effectiveness of e-business adoption and related business outcomes. Methods Figure 2 details of our EBC framework along with associated TOP dimensions which are suited to the three determinants (business strategy, supply chain strategy and e-business adoption) of e-business performance. As illustrated the e-business FM

EM

CM

Business Performance H1

Business Strategy

TI

OI

Supply Chain Strategy

H4

PS

H3

H2

H6

TIn

OIn

E-Business Adoption

H5

SCR

TA

OC

AC

Legend TI OI PS

Technological Infrastructure (IT) Organization Infrastructure Partnership Strategy

TIn OIn SCR

Technological Integration (ERP, EDI) Organisation Integration Supply Chain Relationship

TA OC AC

Technological Adoption Organizational Capability Attitudinal Capability

FM EM CM

Financial Measures Efficiency Measures Coordination Measures

Figure 2. Hypothesized arrangement for the e-business capability framework

JMTM 17,6

812

performance is assesses under financial measures, efficiency measures (or operational measures) and coordination measure related to, respectively, technology, organization and people dimensions. The figure also graphically represents the six hypotheses constructs as outlined in Table II together with dependent variables suited for Structure Equation Modeling procedures. To test the conceptual model in Figure 2 and the associated hypotheses proposed, we conducted a questionnaire survey. To have a broad representation and understanding of the proposed framework, the survey was a stratified sample by industry group (manufacturing; services; IT; finance, insurance and real estate; wholesale and retails trade; others (agriculture, communication, utility services, non-classified). The sites were randomly selected. A predetermined number of targets completes were fifty per industry group. Survey was directed to qualified and experienced individual having a good understanding their businesses. Participants were asked to complete a paper version of the questionnaire comprised of an introductory to outline the purpose and aims of the study. Confidentiality and anonymity were explained on the front page. The set up of the e-mail version (Word document) questionnaires mirrored the paper versions of the survey so that only the mode of completing the survey differed. In a pilot study questionnaires were sent in Microsoft Word format and participants were able to click the value of their choice by using a pull-down menu for each item in each questionnaire. The response rate was 47.7 percent, where 44 percent belong to adopter of e-business group, while 54 percent represent the non-adopter of e-business group. A breakdown of the sample characteristics is illustrated in Table III. Rational of selecting structural equation modelling The structural equation is recognized as a more comprehensive and flexible approach to research design and data analysis than any other single statistical model in standard (Hoyle, 1995). Rather than an exploratory approach SEM takes a confirmatory approach that specifies inter-variable relations a priori, and estimates measurement errors explicitly (Suhr, 1999). However, SEM, require to base the tested model exactly

Hypothesis

Description

H1

Appropriate implementation of business strategy in consideration with TOP dimensions is a significant determinant of perceived business performance The content of supply chain strategy in consideration with TOP dimensions is a significant determinant of perceived business performance Strategic e-business adoption in consideration with TOP dimensions is a significant determinant of perceived business performance Successful e-business implementation is the result of a positive reciprocal effect of business strategy and supply chain strategy in consideration with TOP dimensions Successful of e-business adoption is the result of a positive reciprocal effect of supply chain strategy and e-business adoption strategy in consideration with TOP dimensions Successful e-business implementation is the result of a positive reciprocal effect of business strategy and e-business adoption strategy in consideration with TOP dimensions

H2 H3 H4 H5 Table II. Hypotheses formulation for testing e-business capability model

H6

Country: The United Kingdom Respondents (out of 50)

Industries Manufacturing Services IT Finance, insurance and real estate Wholesale and retails trade Others (agriculture, communication, utility services, non-classified) Total respondent Response rate (percent) Adopter of e-business group Non-adopter of e-business group Job title

25 20 30 20 23 25

E-business capabilities model 813

143 out of 300 47.7 percent 63 80 IS 70 CIO, CTO, VP of IS IS manager, director, planner Other manger in IS department Non IS 73 CEO, president, managing director COO, business operations manager CFO, administration/finance manager Others (IS analyst, marketing VP, other manager)

Keyword CIO VP COO CTO IS

Chief information officer Vice president Chief operating officer Chief technology officer Information system

on theory. Therefore, the use of SEM in a confirmatory way is recommended especially if the purpose is to confirm or reject the proposed hypotheses. The most obvious difference between SEM and other multivariate technique is the use of separate relationships for each of a set of dependent variables (Hair et al., 1998). The SEM becomes a very useful tool when one dependent variable needs to be treated as independent variable in a subsequent analysis. For instance, business strategy, supply chain strategy and e-business adoption factors are treated as initial dependent variables, which in turn become independent variables in terms of their influence on the surveyed companies’ business performance. Other multivariate approaches such as regression analyses are too simplistic and does not allow analysing between independent variables. In comparison of other multivariate method, the SEM applies only the variance/covariance or correlation matrix as its input data. Therefore, focus of SEM is not to understand an individual observation, but on the pattern of relationships across the samples (Hair et al., 1995). In addition, SEM is a comprehensive statistical approach to test hypotheses about relations among observed and latent variables (Hoyle, 1995). The unobserved (latent)

Table III. Demographic characteristics of the survey participants (n ¼ 143)

JMTM 17,6

814

variables is linked to one (variable) that is measurable, thus making its measurement feasible (Bryrne, 2001). Data analysis As mentioned earlier, SEM was used to assess the propositions of the research model using Arbuckle (1997) approach implemented in the SPSS AMOS 4.0 program using maximum likelihood estimation (Bollen, 1989; Byrne, 1998a; Joresko¨g and Sorbo¨m, 1993). Following Marsh et al. (1996, 1988), goodness of fit of the proposed model was computed using comparative fit index (CFI), Tucker-Lewis index (TLI) and root mean square error approximation (RMSEA). In order to further confirm the validation of the model, x 2 test statistic and an evaluation of parameter estimates were calculated. Where various tests of statistical significance and indices of fit aid in the evaluation of a model, we understand that there is always a degree of subjectivity and professional judgment in the selection of a “best” model in multiple group comparison using Marsh et al. (1988). According to Byrne (1998b) and Marsh (1994) there is a well-developed methodology in which the goodness of fit of alternative models are compared. They proposed the least restrictive model that does not require any of the parameter estimates to be the same in different groups, and the most restrictive model that requires all parameter estimates to be the same in the different groups. As part of this study we compared two parallel groups, i.e. of e-business adopters and non-adopters, and tested the invariance of the solution by considering one, any set, or all parameter estimates to be same in two or more groups. Generalizability over groups with tests of invariance The overarching question is whether the EBC model is generalised enough to be replicable in the two groups for two sample groups being examined. One particular interest we have with the analysis is that how the results generalize across the two groups (adopter of e-business and non-adopter of e-business groups) within the sample. We conducted multi-group confirmatory factor analysis (CFAs) and SEMs in which we constrained different parameters (factor loadings, path coefficients, and factor correlations) to be invariant across the 2 groups for the sample (Tables IV and V and Figure 3). We constrained with a set of CFA models to evaluate the invariance of the measurement component of the model (MG2-MG11). Subsequently we had focused specifically on structural equation models (MG12-MG19) to evaluate the appropriate EBC model (Table IV). Multi-group CFAs were conducted from model MG1 to model MG11 considering the factor loadings (First- and second-order) and relevant factor correlations. With the baseline multiple-group model (MG1), no invariance constraints were imposed and parameters for the a-priori model were fit separately to data set for each group. The fit indices were subsequently calculated. With the first (model MG2), only first-order factor loadings are constrained to be equal across the two groups. This model was subsequently used to evaluate the multiple group comparison between adopter and non-adopter of e-business on second-order factor loadings (technological, organisation and people dimensions), correlations among second-order constructs (H1, H2 and H3) and path coefficients (H4, H5 and H6). Results obtained from this model were compared against those based on exclusively non-invariant solution (MG1), provided

DF 761 1,522 1,554 1,559 1,557 1562 1,555 1,555 1,555 1,560 1,560 1,560

x2 871.10 1832.20 1871.21 1869.14 1882.93 1880.17 1876.33 1872.39 1876.62 1872.60 1870.95 1873.28

Total group sample TG1 Multiple group CFA MG1

MG2

MG3

MG4

MG5

MG6 MG7 MG8 MG9 MG10 MG11

1.21 1.20 1.21 1.20 1.20 1.20

1.20

1.21

1.20

1.20

1.20

1.14

x 2/DF

0.90 0.90 0.90 0.90 0.90 0.90

0.90

0.90

0.90

0.90

0.90

0.97

CFI

0.89 0.90 0.89 0.90 0.90 0.90

0.90

0.89

0.90

0.90

0.90

0.97

TLI

0.04 0.04 0.04 0.04 0.04 0.04

0.04

0.04

0.04

0.04

0.04

0.03

RMSEA

CFA INV Uniq CFA INV FL,Uniq CFA INV Uniq CFA INV Uniq CFA INV Uniq CFA INV CFA INV CFA INV CFA INV CFA INV CFA INV ¼ ¼ ¼ ¼ ¼ ¼

1st FL, FCH6; Free ¼ FV, 2nd FL, Uniq 1st FL, FCH4; Free ¼ FV, 2nd FL, Uniq 1st FL, FCH5; Free ¼ FV, 2nd FL, Uniq 1st FL, 2nd FL, FCH6; Free ¼ FV, Uniq 1st FL, 2nd FL, FCH4; Free ¼ FV, Uniq 1st FL, 2nd FL, FCH5; Free ¼ FV, Uniq

¼ 1st FL, 2nd FL, FC(H4-H6), Free ¼ FV,

¼ 1st FL, FC(H4-H6); Free ¼ FV, 2nd FL,

¼ 1st FL, 2nd FL; Free ¼ FV, FC(H4-H6),

¼ 1st FL; Free ¼ FV, FC(H4-H6), 2nd

¼ none; Free ¼ 1st FL, 2nd FL FV, FC,

Full e-commerce capabilities (ECC) model

Model description

Notes: CFI, comparative fitness index; TLI, Tucker-Lewis index; RMSEA, root mean square error of approximation; DF, degrees of freedom; CFA, confirmatory factor analysis; SEM, structural equation model; PM, performance measurement; EBA, e-business adoption; SCS, supply chain strategy; BS, business strategy; 1st FL, factor loading for first-order factors; 2nd FL, factor loadings for second-order factor; FC(H4-H6), factor correlations; FV, factor variances; FC(H4), factor correlation between EBR and BS; FC(H5), factor correlation between SCS and EBA; FC(H6), factor correlation between EBA and BS; PC(H1-H3), path coefficients; PCH1, path coefficient from BS to BP; PCH2, path coefficient from SCS to BP; PCH3, path coefficient from EBA to BP; Uniq, uniqueness. In Model TG1 (see parameter estimates in Table I) the ECC model was fit to the total group, whereas for Models MG1-MG20 the ECC model was fit separately for each of the two groups representing different groups. For Models MG2-MG19, some combination of parameters is required to be invariant across the two groups

Model

E-business capabilities model 815

Table IV. Measurement goodness-of-fit analysis for EBC model fit with respect to the total group and multiple (adopter and non-adopter of e-business) UK group (n ¼ 143)

JMTM 17,6

816

Table V. Structural goodness-of-fit analysis for EBC model fit with respect to the total group and multiple (adopted and non-adopter of e-business) UK group (n ¼ 143)

Model

x2

DF

x 2/DF CFI TLI RMSEA Model description

Multiple group SEM MG12 1883.39 1,565

1.20

0.90 0.90

0.04

MG13 1868.06 1,554

1.20

0.90 0.90

0.04

MG14 1865.42 1,552

1.20

0.90 0.90

0.04

MG15 1871.22 1,554

1.20

0.90 0.90

0.04

MG16 1871.56 1,554

1.20

0.90 0.90

0.04

MG17 1870.02 1,560

1.20

0.90 0.90

0.04

MG18 1869.18 1,560

1.20

0.90 0.90

0.04

MG19 1872.09 1,560

1.20

0.90 0.90

0.04

MG20 1883.39 1,565

1.20

0.90 0.90

0.04

SEM INV ¼ 1st FL, 2nd FL, PC (H1-H3); Free ¼ FV, FC (H4-H6), uniq SEM INV ¼ 1st FL, PC(H1-H3); Free ¼ FV, FC(H4-H6), 2nd FL, uniq SEM INV ¼ 1st FL, PCH3; Free ¼ FV, FC(H4-H6), uniq, 2nd FL SEM INV ¼ 1st FL, PCH1; Free ¼ FV, FC(H4-H6), uniq, 2nd FL SEM INV ¼ 1st FL, PCH2; Free ¼ FV, FC(H4-H6), uniq, 2nd FL SEM INV ¼ 1st FL, 2nd FL, PCH3; Free ¼ FV, FC(H4-H6), uniq SEM INV ¼ 1st FL, 2nd FL, PCH1; Free ¼ FV, FC(H4-H6), uniq SEM INV ¼ 1st FL, 2nd FL, PCH2; Free ¼ FV, FC(H4-H6), uniq SEM INV ¼ 1st FL, 2nd FL, FC(H4-H6), PC; Free ¼ FV, uniq

Notes: CFI, comparative fitness index; TLI, Tucker-Lewis index; RMSEA, root mean square error of approximation; DF, degrees of freedom; CFA, confirmatory factor analysis; SEM, structural equation model; PM, performance measurement; EBR, e-business readiness; SCS, supply chain strategy; BS, business strategy; 1st FL, factor loading for first-order factors; 2nd FL, factor loadings for second-order factors

FM η11

EM η12

β 1,11

β 1,12

CM η 13

β 1,13

Business Performance η1

γ 1,1 φ1,3

Supply Chain Strategy

φ1,2

Business Strateg y

γ1,2 γ1,3 γ1,4 TI η2

OI η3

φ2,3

ξ2

1

Figure 3. Diagram representation for SEM hypothesized model

γ 3,1

γ 2,1

ξ3

γ 2,5 γ 2,6 γ 2,7 PS η4

TIn η5

OIn η6

E-Business Adoption

γ 3,8 γ 3,9 γ 3,10 SCR η7

TA η8

OC η9

AC η 10

that the fit indices were good and differed within a limited range. If good fit indices were obtained from MG2, it supported the appropriateness of the measures across the two groups and satisfied the minimum requirement for subsequent factorial invariance. With model MG3, constraints were imposed on the first- and second-order loadings to measure the difference between the two groups. The path coefficients and correlations with first- and second-order factor loadings were deemed equal across 2 groups. In each of the subsequent CFA models (MG4 – MG11) in Table III, the invariance of the factor loadings was imposed in combination with the invariance of additional sets of parameters on factor correlations and second-order factor loadings. The aim was to assess if the imposition of these added invariance constraints would affect the goodness of fit indices comparing with models MG1 and MG2, respectively. The results thus obtained, therefore support the cross generalizability of the measures and relationship among them across these 2 groups for the sample. Models MG12 to MG19 focus specifically on the structural component of the model – the path coefficients that are critical to tests predictions based on the EBC model (Figure 3). With model MG12, the path coefficients and the first- and second-order factor loadings were required to be the same in each of the two groups, whereas the factor correlations were freely estimated across the both groups. Model MG13 is similar to MG12 except the factor correlations and second-order factor loadings were freely estimated across the groups. The tests of invariance for the three path coefficients (H1-H3) provide a global test that the predicted path coefficients were positive. In order to evaluate models MG14 to MG19, specific path coefficients and factor correlations (H1-H6) were to be freely estimated, while the rest of the factor loadings (first- or second-order depending on model evaluation) showed invariance across the two groups. This was to demonstrate the sensitivity of “goodness of fit” of these models in comparison with model MG2 when certain path coefficients were constraint to test for invariance. In summary, these results illustrate that even the extremely demanding models (with complete invariance of all parameters) provide a good fit to the data (MG12). However, none of these multiple group models stood out clearly as the “best” model. Therefore, further analyses were conducted to evaluate parameter estimates based on the 20 selected models. The two groups were further tested to evaluate: (1) parameters for the second-order factor loadings by constraining first-order factor loadings to be mutually equalled (MG2); and (2) path coefficients and factor correlations by constraining first- and second-order factor loadings to be equal (MG3). Cross-group generalizability: evaluation of parameter estimates between adopter and non-adopter of e-business In order to evaluate parameter estimates between adopter and non-adopter of e-business, we employed the procedures of constructing and evaluating a series of structural equation models across both groups where various model parameters were held invariant across two groups. Models MG1 to MG11 were run to evaluate the multiple groups CFA for adopter and non-adopter of e-business. However, with the baseline multiple-group model (MG1), all of the factor loadings and path correlations

E-business capabilities model 817

JMTM 17,6

818

were feely estimated between the two groups; that provided a good fit of TLI ¼ 0.90, CFI ¼ 0.90, RMSEA ¼ 0.04 and x 2/df ¼ 1.14 (Table IV). In the first test of invariance (model MG2), at first the factor loadings were constrained to be mutually equal across the two groups. Fit indices were found to be very good (TLI ¼ 0.90; CFI ¼ 0.90 and RMSEA ¼ 0.04) and match closely with those based on the totally non-invariant solution (MG1). A good fit of indices result were also produced for model MG3 where first- and second-order factor loadings were held invariant (TLI ¼ 0.92; CFI ¼ 0.93 and RMSEA ¼ 0.04). This supports the appropriateness of the measures across the 2 groups and meets the minimum requirement for factorial invariance (Marsh and Hau, 2003). For each of the subsequent CFA models (MG4 to MG11 in Table IV), the invariance of the factor loadings were imposed in combination with the invariance of additional sets of parameters – first-order factor loadings, second-order factor loading, factor correlations, and uniqueness. Although the imposition of these added invariance constraints resulted in a marginal inferiority of fit, even the highly restrictive models MG4 and MG5 of total invariance (i.e. requiring every parameter to be the same in all two groups) provided a “good fit” to the data that marginally differed with Model MG1 (with no invariance constraints). These results support the cross-cultural generalizability of the measures and the relations among them across these two groups in the sample. Comparison of parameter estimates for hypotheses H1-H6 across two groups based on model MG3 Further analyses were conducted to evaluate the generalizability of the parameter estimates based on model MG3 (Table IV). Of critical importance to this investigation were the three path coefficients (H1-H3) and factor correlations (H4-H6) relating the e-business success to the corresponding e-business capabilities drivers. Subsequently model MG3 was evaluated for comparing path coefficients and factor correlations between adopter and non-adopter groups assuming that all second factors (technology, people and organization elements) invariantly load on business strategy, supply chains strategy and e-business adoption across the two groups. At first, we evaluated our main hypotheses (H1-H3) by having the first- and second-order factor loadings to be invariant (constant) across the two groups. Parameter estimates for this highly restrictive multi-group model MG3 were found to be closely matching with those on the total group model TG1 (Table IV). Results based on model MG3 were concentrated on the path coefficients (H1-H3); and the factor correlations (H4-H6) were the main interests of investigation across the both groups (Table VI). Table VII shows the first-order factor loadings based on model MG2, evaluating the significant affect of TOP dimensions on the three constructs. In addition to providing global support for the EBC model, the invariance of these parameter estimates provided remarkably strong support for the cross-comparison generalizability of prediction based on the EBC model for adopter and non-adopter of e-business. Discussions To understand e-business value and its impact on firm performance, we have empirically tested the proposed EBC framework across two sub-groups for the UK sample. Empirical analysis demonstrated a number of key findings:

H1 H2 H3 Factor correlations H4 H5 H6

Hypotheses 0.29 0.09 0.60 0.28 0.47 0.46

BS ˆ l SCS SCS ˆ l EBA BS ˆ l EBA 0.45 0.09 0.18

0.27 0.45 0.13

Standardized weight, l Adopter Non-adopter

BP ˆ BS BP ˆ SCS BP ˆ EBA

Paths coefficients

0.06 0.07 0.07

0.13 0.14 0.20 0.10 0.05 0.05

0.15 0.15 0.38

Standard error (S.E) Adopter Non-adopter

1.96 2.77 2.86

2.65 0.85 3.90

2.70 0.43 0.94

2.02 3.05 0.84

Critical ratio t-value Adopter Non-adopter

E-business capabilities model 819

Table VI. A comparison between adopter (n ¼ 80) and non-adopter groups (n ¼ 63) based on path coefficients and factor correlations model MG3

JMTM 17,6

820 Table VII. A comparison between adopter (n ¼ 80) and non-adopter groups (n ¼ 63) based on second factor loadings for model MG2

Standardized weight, l Adopter Non-adopter

Paths OI ˆ BS TI ˆ BS PS ˆ BS FM ˆ BP CM ˆ BP EM ˆ BP SCR ˆ SCS OIn ˆ SCS TIn ˆ SCS OC ˆ EBA AC ˆ EBA TA ˆ EBA

0.92 0.87 0.87 0.95 0.95 0.99 0.79 0.79 0.90 0.85 0.75 0.83

0.91 0.96 0.96 0.94 0.91 0.94 0.81 0.88 0.88 0.49 0.63 0.61

Standard error (S.E) Adopter Non-adopter

Critical ratio t-value Adopter Non-adopter

(Fixed) 0.12 0.14 (Fixed) 0.10 0.10 0.17 (Fixed) 0.21 (Fixed) 0.14 0.17

(Fixed) 7.36 6.39 (Fixed) 8.12 6.67 4.51 (Fixed) 5.38 (Fixed) 2.95 4.12

(Fixed) 0.14 0.15 (Fixed) 0.10 0.11 0.15 (Fixed) 0.23 (Fixed) 0.55 0.71

(Fixed) 7.86 7.25 (Fixed) 7.62 6.39 4.51 (Fixed) 5.10 (Fixed) 1.43 1.49

Adopter of e-business From Table VI, sample for adopter of e-business group revealed that business strategy (H1; g11 ¼ 0.31) and e-business adoption (H3; g31 ¼ 0.40) constructs and relevant parameters (paths) are statistically significant with critical ratios of 2.65 and 3.90, respectively, (critical ratio . 1.96 is the nominal requirement of significance). This clearly provides evidence that these constructs have a positive and significant impact on the overall business performance. However, there was no significant confirmation linking “supply chain strategy to business performance” for group of e-business adopter with a non-significant standardize value of (H2; g21 ¼ 0.12) and critical ratio of 1.10. All constructs belonging to business strategy, supply chain strategy and e-business adoption strategy had strong significant reciprocal effect for the e-business adopter companies. As shown in Figure 4, supply chain strategy and e-business adoption (H5) are the highest correlating factors with value f23 ¼ 0.47 and c.r. ¼ 2.77. Correlation Business Performance H1

0.29** 0.27**

H6

H2

0.09 0.45**

H3

0.60** 0.13

0.46** 0.18 Business Strategy

0.28* 0.45**

Supply Chain Strategy

H4

Figure 4. Empirical results – split group (estimate on the sub-groups of adopter and non-adopter of e-business)

Bold : Adopter of E - Business Group (n = 80) Non Bold : Non Adopter of E - Business Group (n = 63) Italic : Not Significant path * p < 0.05 ** p < 0.01

0.47** 0.09 H5

E-Business Adoption

between business strategy and e-business adoption (H6) revealed the second highest standardised value f13 ¼ 0.46 and c.r. ¼ 2.86, followed by correlation between business strategy and supply chain strategy construct (H4) with f12 ¼ 0.28 with c.r. ¼ 1.96. E-business adoption is evaluated as a significant and strong determinant of “business performance” in comparison with other constructs, and postulates a major reason towards a successful e-business adoption in the UK companies. “Attitudinal capability (AC)” for first-order e-business adoption construct revealed a strong and significant standardised value g310 ¼ 0.75 with c.r. ¼ 2.95. This supports the academic view that successful e-business adoption require a dedicated individual (usually the Chief Executive Office (CEO) paying attention to a multitude of good management practices to develop right attitudes for his/her employees to adopt organisational change (Tidd et al., 2001). “Technology adoption (TA)” and “organisation capability (OC)” also observed to be a significant and strong determination of e-business implementation for adopter of e-business in sample with g38 ¼ 0.83: c.r. ¼ 4.12; g39 ¼ 0.75: c.r. ¼ fixed. Adopter of the successful e-business in the UK also takes into consideration the importance of having a well defined supply chain strategy. However, supply chain strategy construct did not directly contribute towards the business performance. Some of the reasons may due to the misalignment supply chain incentive that characterizes the lack of consistent incentives among supply chain partners such as objectives among the supply chain partners and lack of shared visions (and risks) between the supply chain partners (Piplani and Fu, 2005). This is further support by Koh and Simpson (2005) article which stated that: . . . the new economy entails e-business and knowledge-driven enterprises that could lead to more responsive and agile methods to deal with change and uncertainty in dynamic manufacturing environments.

They further stated the importance of responsiveness and agility especially for manufacturing enterprises to achieve key competitive advantages. Nevertheless, the results in Figure 4 show that there is a strong correlation between e-business adoption and supply chain strategy. This implies that adopters of e-business in UK may already been on a supply chain integration maturity stage before embarking on the e-business journey. The theoretical model illustrated in this paper has identified and confirmed that a successful e-business adoption require a comprehensive business strategy along with supply chain and e-business strategy; developed on the embedded e-technology as well as considering organisational and attitudinal factors as identified by Stevens (1989). It also validates our claims that in order to adopt a successful e-business, it is vital to assess e-business readiness for the firm. Where some previous studies have focused on which dimensions are important in supply chain strategies and business strategies (Gattorna, 2002), our model measures the influence of e-business adoption in order to become a successful e-business firm. The challenge for the companies is to move up to a whole new level of performance – the second wave – on the back of e-business – e-procurement, e-fulfilment, and shared services that will lead to Internet-enabled supply chain (Gattorna, 2002).

E-business capabilities model 821

JMTM 17,6

822

Non-adopter of e-business As for as the non-adopter of e-business group, only two path coefficients (H1 and H2) provide positive and significant results. Standardize path coefficient business strategy construct (H1; g11 ¼ 0.27, c.r. ¼ 2.02) and supply chain strategy construct (H2; g21 ¼ 0.45, c.r. ¼ 3.05) provide strong link to the business performance. However, as suspected, e-business adoption construct (H3; g31 ¼ 0.13, c.r. ¼ 0.84) shows a positive but insignificant value, which means that e-business adoption strategy employed by the non-adopters is either non-existent or negatively contribute towards the business performance. However, the analysis validates fully the conceptual and theoretical construct of the e-business model. We also studied the factor correlation results for the both groups. In comparison with adopter of e-business group, as expected, only correlation between supply chain strategy and business strategy (H4) provide a positive correlation with standardise value of f12 ¼ 0.45 and c.r. ¼ 2.70. Meanwhile, a low and non-significant correlation are recorded between supply chain strategy and e-business adoption (H5; f23 ¼ 0.09 and c.r. ¼ 0.43) and business strategy and e-business adoption (H6; f13 ¼ 0.18 and c.r. ¼ 0.94). There were no significant causal paths found linking “TA, OC and AC to e-business adoption” construct (Table VII). Perhaps for this group the main barrier to e-business adoption could be unwillingness of mangers to be responsible for technological change (Kalakota and Robinson, 2001), complexity of available e-commerce services (Bodorick et al., 2002) and lack of required skills and knowledge (Lawson et al., 2003). It is not surprising that the business strategy and supply chain capabilities are the main contributor for business performance for non-adopter of e-business. It is apparent that for organisation to be successful, SCM need to be given a higher level of strategic importance (Maede, 1998; Philip and Pedersen, 1997). The result also support the view that organizations who articulate their strategic objectives and plans relating to SCM are likely to perceive business benefits from traditional brick and mortar businesses (Figure 5). Business Performance

Significant Not - Significant

2

Business Strategy

Figure 5. Summary of findings for non-adopter of e-business describing strength of business capability constructs with business performance

1

3 Supply Chain Strategy

E-Business Adoption

Managerial implications This study offers several implications for academics and practitioners. Firstly, the proposed conceptual model would be able to provide an efficient framework to assess the firm’s readiness for Internet adoption in hope to reap the e-business benefits. This EBC framework has included a number of e-business requirements that need to be taken into consideration within the firm, and includes business strategy, supply chain and e-business adoption strategy which each of the factor describe the needs of considering “technology” “organization” and “technology” dimensions. These specific factors of indicator will be able to measure the readiness of firm for emerging e-business. In addition, these indicators also allow manages to identify which of the factors that is lack of strategic implementation when considering e-business adoption. Therefore, managers will be able to evaluate the readiness for current and future e-business development within their firms and how they must enhance “technology” “organization” and “technology” dimensions witning each of the EBC factors to improve e-business performance. Secondly, the results also suggest that firms belong non-adopter of e-business group must pay attention to their technological, organisational, and human capability for improving e-business performance. These capabilities are critical when firms are planning or at the very initial stage of e-business adoption, where most processes are at low integration level and full of manual work (Hsin and Shaw, 2005). By doing top managers will be encourage to initiate in developing a financial and human plan to accommodate the appropriate resources and to handle the complexity of IT infrastructure. Firms that intend to venture into e-business would need to consider on issues such as IT promotion and training in order to overcome the “organisation” barriers by offering training and knowledge for system integration, standards development, and process automation as well as to overcome possible IT resistance (Hsin and Shaw, 2005). Thirdly, the results obtain from adopter of e-business groups from the UK indicated that the significant drives for adopting e-business in their firms contributed much to business partner’s willingness (“people dimension”), technology capability and empowerment (“organisational dimension”). The results would suggest that in order to improve supply chain readiness for e-business management maybe find the needs to introduce support programs to increase partner willingness and offering initiative such as training, on-site assistance, and financial resources to improve partner capability (Barua et al., 2002; Hsin-Lu and Shin-Horng, 2005). Such initiatives in combination with suitable market power enable firms to have a higher chance of e-business success. The result also highlighted the crucial role of collaboration readiness as the firm starts to implement more advanced e-business IT. Conclusion This study has tested a conceptual model that advanced the theoretical basis of the technology- organization-people framework. Our results have demonstrated the usefulness of this conceptual model by identifying factors affecting business performance of a firm. We have developed several multi-item constructs, including e-business adoption, business strategy and SCM strategy. These instruments have passed various reliability and validity tests, and could be used in future studies. Third, grounded in theory and empirical data, we have demonstrated varying relationships

E-business capabilities model 823

JMTM 17,6

824

between the technology – organization – people factors and e-business value. These associations may useful for other researchers to develop their own models and hypotheses (Table VIII). Firms that positioned themselves in the “new economy” by embracing Internet technology will be expected to have a greater degree of success than traditional firms (Leadbeater, 2000). This paper has promoted the need to understand how internal (“technology” and “organization”) and external (“people”) factors encompass in business strategy, supply chain strategy affecting on e-business adoption and development. The theoretical model illustrated in this paper has identified and confirmed that a successful e-business adoption require a comprehensive business strategy along with supply chain and e-business strategy; developed on the embedded e-technology as well as considering organisational and attitudinal factors as identified by Stevens (1989). Results show that the adopter of the successful e-business in the UK takes into consideration the importance of having a well-defined supply chain strategy. However, supply chain strategy construct did not directly contribute towards the business performance. Comparisons using the structural model indicate non-adopter of e-business lack appropriate e-business strategy for a successful e-business implementation in their companies. The research method used in this paper has its merits and limitations. The first strength lied with the use of SEM where by this technique assess the theoretical and measurement models. This approach requires a whole set of validation rules to be applied and thereby provide a greater confidence in the results and conclusions (Croteau et al., 2001). The measurement model is tested for unidimensionality, reliability, convergent and

Hypothesis

Description

H1

Appropriate implementation of business strategy in consideration with TOP dimensions is a significant determinant of perceived business performance The content of supply chain strategy in consideration with TOP dimensions is a significant determinant of perceived business performance Strategic e-business adoption in consideration with TOP dimensions is a significant determinant of perceived business performance Successful e-business implementation is the result of a positive reciprocal effect of business strategy and supply chain strategy in consideration with TOP dimensions Successful of e-business adoption is the result of a positive reciprocal effect of supply chain strategy and e-business adoption strategy in consideration with TOP dimensions Successful e-business implementation is the result of a positive reciprocal effect of business strategy and e-business adoption strategy in consideration with TOP dimensions

H2

H3 H4

H5

Table VIII. Hypotheses result for the e-business adopter and non-adopter groups

H6

Sub-sample groups Adopter

Non-adopter

Supported

Supported

Not supported

Supported

Supported

Not supported

Supported

Supported

Supported

Not supported

Supported

Not supported

discriminant validity. From methodological point of view SPSS AMOS 4.0, compared with the other SEM software, is a powerful tool that eases the statistical analysis of the four second-order constructs (whereby each consists of three first-order constructs, respectively). The use of the survey approach facilitates a greater number of variables compare with other methods such as case study or experiments (Galliers, 1985). References Arbuckle, J. (1997), AMOS Users’ Guide Version 3.6. Smallwaters Corporation. Barratt, M. (2002), “Understand the meaning of collaboration in the supply chain”, Supply Chain Management: An International Journal, Vol. 9 No. 1, pp. 30-42. Barratt, M.A. and Green, M. (2001), “The cultural shift; the need for collaborative culture”, Conference Proceeding of Supply chain Knowledge 2001, Canfield School of Management, November. Barua, A. and Mukhopadhyay, T. (2000), “Information technology and business performance: past, present and future”, in Zmud, R.W. (Ed.), Framing the Domains of IT Management: Projecting the Future through the Past, Pinnaflex Education Resources, Cincinnati, OH, pp. 65-84. Barua, A., Konana, P., Whinston, A.B. and Yin, F. (2002), “Managing e-business transformation: opportunities and value assessment”, Sloan Management Review, available at:www. utexas.edu(accessed June 19, 2002). Bodorick, P., Dhaliwal, J. and Jutla, D. (2002), “Supporting the e-business readiness of small and medium-sized enterprises: approaches and metrics”, Internet Research: Electronic Networking Applications and Policy, Vol. 12 No. 2, pp. 139-64. Bollen, K.A. (1989), Structural Equations with Latent Variables, Wiley, New York, NY. Brynjolfsson, E. and Kahin, B. (2000), Understanding the Digital Economy, MIT Press, Cambridge, MA, October. Byrne, B.M. (1998a), Structural Equation Modeling with AMOS: Basic Concepts, Applications, and Programming, Lawrence Erlbaum, Mahwah, NJ. Byrne, B.M. (1998b), Structural Equation Modeling with LISREL, PRELIS, and SIMPLIS: Basic Concepts, Applications and Programming, Lawrence Erlbaum, Mahwah, NJ. Coviello, N. and McAuley, A. (1999), “Internationalisation and the smaller firm: a review of contemporary empirical research”, Management International Review, Vol. 39 No. 3, pp. 223-40. Croteau, A., Simona, S., Raymond, L. and Bergeron, F. (2001), “Organizational and technological infrastructures alignment”, Proceedings of the 34th Hawaii International Conference on System Sciences. Damien, P. (2005), “Business to business e-commerce implementation and performance”, Supply Chain Management: An International Journal, Vol. 10 No. 2, pp. 96-113. Daniel, E. (2003), “Exploration of the inside-out mode: e-commerce integration in SEMs”, Journal of Small Business and Development, Vol. 10 No. 3, pp. 223-49. Das, S.R., Zhara, S.A. and Warkentin, M.E. (1991), “Integrating the content and process of strategic MIS planning with competitive strategy”, Decision Sciences, pp. 953-84. DTI (2002), “The government’s expenditure plans 2001-02 to 2003-04, department of trade and industry, London”, available at: www.dti.gov.uk/expenditureplan/expenditure2001/ objectivea/chapter2/section2.htm (accessed 19 November). Earl, M.J. (2000), “Evolving the e-business”, Business Strategy Review, Vol. 11 No. 2, pp. 33-8.

E-business capabilities model 825

JMTM 17,6

826

Fawcett, S.E. and Magnan, G.M. (2001), Achieving World-Class Supply Chain Alignment: Benefits, Barriers and Bridges, Centre for Advance Purchasing Studies, Tempe, AZ. Galliers, R.D. (1985), “In search of a paradigm for information systems research”, in Munford, E. et al. (Eds), Research Methods in Information Systems, North-Holland, Amsterdam, pp. 281-97. Gattorna, J.L. (2002), “The e-supply chain reaches Asian shores”, Ascet, Vol. 2, Accenture. Haapaniemi, P., Huler, S., McDermott, M. and Benchley, R. (2000), “Tuning up for the e-commerce fast lane”, Chief Executive, Leadership in the New Economy, Supplement, August, pp. 20-5. Haig, M. (2002), “The e-commerce report”, Internet Works, December, p. 8. Hau, L. (2000), “Creating value through supply chain integration”, Supply Chain Management Review, September/October, pp. 30-6. Hsin, L.C. and Shaw, M.J. (2005), “A roadmap to adopting emerging technology in e-business – an empirical study”, Management, Vol. 8 No. 2, pp. 31-47. Hsin-Lu, C. and Shin-Horng, C. (2005), “Assessing the readiness of internet-based IOS and evaluating its impact on adoption”, Proceedings of the 38th Annual Hawaii International Conference on System Sciences (HICSS05) – Track 7, p. 207b. Ian, F., Ulf, J. and Beverly, W. (2004), “Factors impacting on e-business adoption and development in the smaller firm”, International Journal of Entrepreneurial Behaviour & Research, Vol. 10 No. 3, pp. 178-91. Ireland, R. and Bruce, R. (2000), “CPFR: only the beginning of collaboration”, Supply Chain Management Review, September/October, pp. 80-8. Joresko¨g, K.G. and Sorbo¨m, D. (1993), LISREL 8: Structural Equation Modelling with the SIMPLIS Command Language, Scientific Software International, Chicago, IL. Kalakota, R. and Robinson, M. (2001), E-business 2.0: Road Map for Success, Addison-Wesley, Harlow. Kaplan, R.S. and Norton, D.P. (2004), “Measuring the strategic readiness of intangible assets”, Harvard Business Review, Vol. 82 No. 2, p. 52, 12p, 4c; (AN 12108631). Keeling, K., Vassilopoulou, K., McGoldrick, P. and Macaulay, L. (2000), “Market realities and innovation in small to medium enterprises: facilitators and barriers to the use of electronic commerce”, New Product Development and Innovation Management, March/April, pp. 57-70. Keogh, W., Jack, S.L., Bower, J. and Crabtree, E. (1998), “Small, technology-based firms in the UK oil and gas industry: innovation and internationalisation strategies”, International Small Business Journal, Vol. 17 No. 1, pp. 57-68. Keoy, K.H., Hafeez, K. and Drake, R. (2002), “Supply chain management strategies for e-commerce companies, ICT research & application innovation at work”, paper presented at Conference on IT Research and Applications, Malaysia. Koh, S.C.L. and Simpson, M. (2005), “Coordination framework for supply chain inventory alignment”, Journal of Manufacturing Technology Management, Vol. 16 No. 6. Lawson, R., Alcock, C., Cooper, J. and Burgess, L. (2003), “Factors affecting adoption of electronic commerce technologies by SMEs: an Australian study”, Journal of Small Business and Enterprise Development, Vol. 10 No. 3, pp. 265-76. Leadbeater, C. (2000), Living on Thin Air: The New Economy, Penguin, London. Lee, C.S. and Kalakota, R. (2001), “An analytical framework for evaluating e-commerce business models and strategies”, Internet Research: Electronic Networking Applications and Policy, Vol. 11 No. 4, pp. 349-59.

Levitt, T. (1983), “The globalisation of market”, Harvard Business Review, May/June. Maede, L. (1998), “Strategic analysis of logistics & supply chain management systems using the analytical network process”, Transportation Research Part E – Logistics & Transportation Review, Vol. 34 No. 3, pp. 201-15. Marsh, H.W. (1994), “Confirmatory factor analysis models of factorial invariance: a multifaceted approach”, Structural Equation Modelling, Vol. 1, pp. 5-34. Marsh, H.W. and Hau, K. (2003), “Relations between academic self-concept and achievement in mathematics and language: cross-cultural generalisability”, Self-concept Enhancement and Learning Facilitation Research Centre University of Western Sydney, Australia paper presented at NZARE AARE, Auckland. Marsh, H.W., Balla, J.R. and McDonald, R.P. (1988), “Goodness of fit indexes in confirmatory factor analysis: the effect of sample size”, Psychological Bulletin, Vol. 103, pp. 391-410. Marsh, H.W., Balla, J.R. and Hau, K.T. (1996), “An evaluation of incremental fit indices: a clarification of mathematical and empirical processes”, in Marcoulides, G.A. and Schumacker, R.E. (Eds), Advanced Structural Equation Modelling Techniques, Erlbaum, Hillsdale, NJ, pp. 315-53. Martinsons, M.G. and Martinsons, V. (2002), “Rethinking the value of IT, again”, Communications of the ACM, Vol. 45 No. 7, pp. 25-6. Melymuka, K. (2000), “Survey finds companies lack e-commerce blueprint”, Computerworld, Vol. 34 No. 16, pp. 38-42. Morton, M.S. (1991), The Corporation of the 1990s: Information Technology and Organizational Transformation, Oxford University Press, New York, NY. Oftel (2002), “Business internet usage in the UK”, February, available at: www.oftel.gov.uk (assessed January 2004). Ontario (2001), “The wisdom exchange e-business readiness assessment”, available at: www. ontariocanada.com/ontcan/en/PDF_HTML/Priority-2/w-e-e-business_web.htm Philip, G. and Pedersen, P. (1997), “Inter-organisation information systems: are organisations in Ireland deriving strategic benefits from internet?”, International Journal of Information Management, Vol. 23, pp. 201-21. Piplani, R. and Fu, Y. (2005), “Coordination framework for supply chain inventory alignment”, Journal of Manufacturing Technology, Vol. 16 No. 6. Poon, S. and Swatman, P.M.C. (1999), “An exploratory study of small business internet commerce issues”, Information & Management, Vol. 35, pp. 9-18. Porter, M.E. (1985), Competitive Advantage, Free Press, New York, NY, pp. 33-57. Porter, M. (2001), “Strategy and the internet”, Harvard Business Review, Vol. 79 No. 3, p. 62. Rosenzweig, E.D., Roth, A.V. and Dean, J.W. (2003), “The influence of integration strategy on integration strategy on competitive capabilities and business performance: an exploratory study of consumer products manufacturer”, Journal of Operations Management, Vol. 21 No. 4, pp. 437-56. Ross, J., Vitale, M. and Weill, P. (2001), “From place to space: migrating to profitable electronic commerce business models”, Working Paper No. 324, MIT Sloan School of Management, No. 4358-01. Sabath, R. and Fontanella, J. (2002), “The unfulfilled promise of supply chain collaboration”, Supply Chain Management Review, July/August, pp. 24-9. Simpson, M. and Docherty, A.J. (2004), “E-commerce adopter support and advice for UK SMEs”, Journal of Small Business and Enterprise Development, Vol. 11 No. 3, pp. 315-28.

E-business capabilities model 827

JMTM 17,6

828

Stevens, G. (1989), “Integrating the supply chain”, International Journal of Physical Distribution & Materials Management, Vol. 19 No. 8, pp. 3-8. Storey, D.J. (1994), Understanding the Small Business Sector, Routledge, London. Straub, D. (1989), “Validating instruments in MIS research”, MIS Quarterly, Vol. 13 No. 2, pp. 147-69. UK Online (2002), UK Online Annual Report, Department of Trade and Industry, London. Vickery, S.K., Jayaram, J., Droge, C. and Calantone, R. (2003), “The effects of integrative supply chain strategy on customer service and financial performance: an analysis of direct versus indirect relationships”, Journal of Operations Management, Vol. 21 No. 5, pp. 523-39. Von Hoffman, C. (2001), “Built to move”, CIO – Chief Information Officer, Vol. 13 No. 16, pp. 120-8. Webb, B. and Sayer, R. (1998), “Benchmarking small companies on the internet”, Long Range Planning, Vol. 31 No. 6, pp. 815-27. Wheeler, B.C. (2002), “NEBIC: a dynamic capabilities theory for assessing Net-enablement”, Information Systems Research, Vol. 13 No. 2, pp. 125-46. Zhu, K., Kraemer, K.L. and Xu, S. (2003), “E-business adoption by European firms: a cross country assessment of the facilitators and inhibitors”, European Journal of Information Systems, Vol. 12 No. 4, pp. 251-68. Zhu, K., Ken, K., Sean, X. and Jason, D. (2004), “Information technology payoff in e-business environments: an international perspective on value creation of e-business in the financial services industry”, Journal of Management Information Systems (JMIS), Vol. 21 No. 1, pp. 17-54. Further reading Austin, J.E. (1990), Managing in Developing Countries, Free Press, New York, NY. Dewan, S. and Kraemer, K.L. (2000), “Information technology and productivity: evidence from country-level data”, Management Science, Vol. 46 No. 4, pp. 548-62. Jarvenpaa, S. and Leidner, D. (1998), “An information company in Mexico: extending the resource-based view of the firm to a developing country context”, Information Systems Research, Vol. 9 No. 4, pp. 342-61. UK Online for Business (2002), “Majority of small businesses developing an e-strategy according to latest IM research”, Online 15th March, available at: www.ukonlineforbusiness.gov.uk/ main/news/mainnewsviewer.jsp?contentID ¼ 1061 Watson, R.T., Kelly, G., Galliers, R.D. and Brancheau, J.C. (1997), “Key issues in information systems management: an international perspective”, Journal of Management Information Systems, Vol. 13 No. 4, pp. 91-115. Zhu, K., Kenneth, K., Sean, X. and Jason, D. (2004), “Information technology payoff in e-business environments: an international perspective on value creation of e-business in the financial services industry”, Journal of Management Information Systems, Vol. 21 No. 1, pp. 17-56. Corresponding author Kay Hooi Keoy can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

The current issue and full text archive of this journal is available at www.emeraldinsight.com/1741-038X.htm

An analysis of the relationship between TQM implementation and organizational performance

TQM and organizational performance

Evidence from Turkish SMEs

Received August 2005 Revised January 2006 Accepted February 2006

Mehmet Demirbag

829

Management School, University of Sheffield, Sheffield, UK

Ekrem Tatoglu School of Business Administration, Bahcesehir University, Istanbul, Turkey

Mehmet Tekinkus Faculty of Economics and Administrative Sciences, Gaziantep University, Gaziantep, Turkey, and

Selim Zaim Faculty of Economics and Administrative Sciences, Fatih University, Istanbul, Turkey Abstract Purpose – The principal aim of this paper is to determine the critical factors of total quality management (TQM) and to measure their effect on organizational performance of SMEs operating in Turkish textile industry. Design/methodology/approach – Data for this study was collected using a self-administered questionnaire that was distributed to 500 SMEs in textile industry in the city of Istanbul in Turkey selected randomly from the database of Turkish Small Business Administration (KOSGEB). Of the 500 questionnaires posted, a total of 163 questionnaires were returned. Findings – Using exploratory and confirmatory factor analyses, seven empirically validated dimensions of TQM were identified. The structural equation modelling technique was employed to investigate the relationship between the implementation of TQM practices and organizational performance. Data analysis reveals that there is a strong positive relationship between TQM practices and non-financial performance of SMEs, while there is only weak influence of TQM practices on financial performance of SMEs. With only a mediating effect of non-financial performance that the TQM practices has a strong positive impact on financial performance of SMEs. Research limitations/implications – The sample is restricted to only a single region and a single industry, so it would be strongly recommended that data be gathered from various parts of Turkey including both various manufacturing and service industries. As the data in this study were collected from top managers of organizations on the basis of their subjective evaluations, objective performance indicators should also be employed in the analysis. Originality/value – Despite some attempts on the applicability of TQM practices and advanced manufacturing technologies as well as their impact on organizational performance of SMEs, there is a

Journal of Manufacturing Technology Management Vol. 17 No. 6, 2006 pp. 829-847 q Emerald Group Publishing Limited 1741-038X DOI 10.1108/17410380610678828

JMTM 17,6

830

lack of systematic empirical evidence regarding the extent of TQM implementation and its effect on performance of SMEs in emerging market economies. This paper presents new data and empirical insights into the relationship between TQM implementation and organizational performance in SMEs operating in Turkey. Keywords Total quality management, Organizational performance, Small to medium-sized enterprises, Turkey Paper type Research paper

Introduction Quality has become one of the most important drivers of the global competition today. Intensifying global competition and increasing demand for better quality by customers have caused more and more companies to realize that they will have to provide high quality product and/or services in order to successfully compete in the marketplace. To meet the challenge of this global competition, many businesses have invested substantial resources in adapting and implementing total quality management (TQM) strategies. TQM can be defined as a holistic management philosophy aiming at continuous improvement in all functions of an organization to produce and deliver commodities or services in line with customers’ needs or requirements by better, cheaper, faster, safer, easier processing than competitors with the participation of all employees under the leadership of top management. The role of TQM is widely recognized as being a critical determinant in the success and survival of both manufacturing and service organizations in today’s competitive environment. TQM is also seen as a source of competitive advantage (Powel, 1995; Hackman and Wageman, 1995; Douglas and Judge, 2001), innovation (Singh and Smith, 2004), change and new organizational culture (Irani et al., 2004). Any decline in customer satisfaction due to poor service quality would be a serious cause of organizational failure. Consumers are becoming increasingly aware of rising standards in product/service quality, prompted by competitive trends, which have developed higher expectations. Despite some attempts on the applicability of TQM practices and advanced manufacturing technologies as well as their impact on organizational performance of SMEs (Ahire and Golhar, 1996; McAdam and McKeown, 1999; Yusof and Aspinwall, 2000; Cagliano et al., 2001; Sun and Cheng, 2002; Lee, 1998, 2004; Raymond, 2005; Dangayach and Desmukh, 2005), there is a lack of systematic empirical evidence regarding the extent of TQM implementation and its effect on performance of SMEs in emerging market economies such as Turkey. SMEs play a very crucial role to the economies of most emerging nations from the viewpoint of generating employment and economic growth. They account for more than half of the employment and added value in most countries (UNCTAD, 1993). Similar trend is also observed in Turkey where SMEs constitute 99 per cent of all business establishments and employ 53 per cent of the workforce in the manufacturing sector (Taymaz, 1997). In view of the fact that the success of small business has a direct impact on the national economy, this paper presents new data and empirical insights into the relationship between TQM implementation and organizational performance in SMEs operating in Turkey. The paper is organized as follows: The next section provides a review of the theoretical literature and sets out the hypotheses of the study. The research methods are presented in the section Research methodology. The next section presents the results followed by conclusion and managerial implications.

Literature review and hypotheses development Although the literature on TQM includes a rich spectrum of works, there is no consensus on the definition of quality. The notion of quality has been defined in different ways by different authors. Gurus of the TQM practices such as Garvin, Juran, Crosby, Deming, Ishikawa and Feigenbaum all provided their own definitions of quality concept and TQM. Garvin (1987) proposed a definition of quality in terms of the transcendent, product based, user based, and manufacturing and value-based approaches. He also identified eight attributes to measure product quality (Garvin, 1987). Juran defined quality as “fitness for use” and focused on a trilogy of quality planning, quality control, and quality improvement (Mitra, 1987). Similarly, Crosby (1996) defined quality as “conformance to requirements or specifications” that is based on customer needs. He identified 14 steps for a zero defect quality improvement plan to achieve performance improvement. According to Deming, quality is a predictable degree of uniformity and dependability, at low cost and suited to the market. He also identified 14 principles of quality management to improve productivity and performance of the organization (Deming, 1986). Ishikawa also emphasized importance of total quality control to improve organizations’ performance. He contributed to the quality literature by introducing a cause and effect diagram (Ishikawa diagram) to diagnose quality problems (Mitra, 1987). In a similar vein, Feigenbaum introduced the concept of organization-wide total quality control and defined quality as “the total composite product and service characteristics of marketing, engineering, manufacturing and maintenance through which the product and service in use will meet the expectations by the customer” (Mitra, 1987). Major common denominators of these quality improvement plans include management commitment, strategic approach to a quality system, quality measurement, process improvement, education and training, and eliminating the causes of problems. TQM is the culture of an organization committed to customer satisfaction through continuous improvement. This culture varies from one country to another and between different industries, but has certain essential principles, which can be implemented to secure greater market share, increased profits, and reduced costs (Kanji and Wallace, 2000). Management awareness of the importance of TQM, alongside business process reengineering and other continuous improvement techniques was stimulated by the benchmarking movement to seek, study, implement and improve on best practices (Zairi and Ahmed, 1999). A review of extant literature on TQM and continuous improvement programs identifies 12 common aspects: Committed leadership, adoption and communication of TQM, closer customer relationships, benchmarking, increased training, open organization, employee empowerment, zero defects mentality, flexible manufacturing, process improvement, and measurement. Furthermore, to determine critical factors of TQM, various studies were undertaken and different instruments were developed by individual researchers and institutions such as Malcolm Baldrige Award, EFQM (European Foundation for Quality Management), and the Deming Prize criteria. Based on these studies, a wide range of management issues, techniques, approaches, and systematic empirical investigations have been generated. Saraph et al. (1989) developed 78 items related to TQM practices, which were classified into eight critical factors to measure the performance of TQM in an organization. They labelled these critical factors as: Role of divisional top management and quality policy, role

TQM and organizational performance 831

JMTM 17,6

832

of the quality department, training, product and service design, supplier quality management, process management, quality data and reporting, and employee relations. Flyyn et al. (1994) developed another instrument which they identified seven quality factors of TQM. These are top management support, quality information, process management, product design, workforce management, supplier involvement, and customer involvement. This instrument is in close resemblance to the preceding instrument developed by Saraph et al. (1989). In a later study, Flyyn et al. (1995) measured the impact of TQM practices on quality performance and competitive advantage. On the other hand, Anderson et al. (1994) developed the theoretical foundation of quality management practice by examining Deming’s 14 points. They reduced the number of factors from 37 to 7 using the Delphi method, which consist of visionary leadership, internal and external cooperation, learning, process management, continuous improvement, employee fulfilment and customer satisfaction. In a similar vein, using the Malcolm Baldrige Award criteria Black and Porter (1996) identified ten empirically validated critical TQM factors, which include corporate quality culture, strategic quality management, quality improvement measurement systems, people and customer management, operational quality planning, external interface management, supplier partnerships, teamwork structures, customer satisfaction orientation, and communication of improvement information. In addition to Black and Porter (1996), various authors also assessed the validity of Malcolm Baldrige Award criteria (Wilson and Collier, 2000; Flynn and Saladin, 2001). Ahire et al. (1996) developed 12 integrated quality management constructs, which were labelled as supplier quality management, supplier performance, customer focus, statistical process control usage, benchmarking, internal quality information usage, employee involvement, employee training, design quality management, employee empowerment, product quality, and top management commitment. Performance measurement is very important for the effective management of an organization. According to Deming without measuring something, it is impossible to improve it. Therefore, to improve organizational performance, one needs to determine the extent of TQM implementation and measure its impact on business performance (Madu et al., 1996; Gadenne and Sharma, 2002). Traditionally, organizational performance has been measured by using financial indicators, which may include inter alia profit, market share, earnings, and growth rate. Kaplan and Norton (1996) emphasized that financial indicators would measure only past performance. Therefore, in order to overcome potential shortcomings of traditional organizational performance systems they added non-financial categories to the traditional performance measurement system. There is a relatively large body of empirical studies that measure business performance by TQM criteria (Samson and Terziovski, 1999; Flyyn et al., 1995; Wilson and Collier, 2000; Fynes and Voss, 2001; Flynn and Saladin, 2001; Montes et al., 2003; Benson et al., 1991; Choi and Eboch, 1998). These studies explore a variety of theoretical and empirical issues. If TQM plan is implemented properly, it produces impact on a wide range of areas including understanding customers’ needs, improved customer satisfaction, improved internal communication, better problem solving and fewer errors. In the following subsections, we develop a number of hypotheses to investigate the relationship between the implementation of TQM practices and organizational performance in SMEs. First, we examine the relationship between the critical factors of

TQM and their effect on both financial performance and non-financial performance. Next, we investigate to what extent non-financial performance mediates the relationship between TQM practices and financial performance. TQM practices and financial performance Empirical studies investigating the relationship between TQM practices and financial performance have produced mixed results. These studies either use stock price performance (Hendricks and Singhal, 1996, 2001; Easton and Jarrel, 1998) or perceptual measures developed by researchers themselves (Powel, 1995; Kaynak, 2003; Samson and Terziovski, 1999; Prajogo and Sohal, 2006). Hendricks and Singhal (1996) studied award-winning companies (as a proxy for TQM implementation) to establish a link between TQM and stock price performance but found no evidence of long-term abnormal performance. In contrast to the findings of Hendricks and Singhal (1996), Easton and Jarrel (1998) found significant relationship between stock-price performance and TQM implementation. A follow up study by Hendricks and Singhal (2001) with a larger dataset revealed that in the post implementation period, the sample of effective TQM implementers significantly outperformed the various matched control groups. Douglas and Judge (2001) used perceptual measures of financial performance (alongside with expert rated performance measures). Their results indicated that the level of TQM implementation was positively and significantly related to both perceived financial performance of a hospital and its industry-expert rated performance. It appears that the degree to which the entire TQM philosophy is implemented strongly correlated with financial performance perception (Kaynak, 2003). When the firm size is taken into account, evidence seems to get mixed. Some TQM advocates argue that TQM cannot produce consistent financial performance for SMEs (Schmidt and Finnigan, 1992; Powel, 1995; Strubering and Klaus, 1997), while others found some significant results in TQM implementations in SMEs (Ahire and Golhar, 1996; Hendricks and Singhal, 2001). Hendricks and Singhal’s (2001, p. 287) analyses indicate that smaller firms tend to benefit more from TQM as compared to larger firms. This finding contradicts with some of the earlier arguments that TQM is less beneficial to smaller firms. While in general the evidence seems conflicting at least for SMEs, we expect that: H1. TQM practices have a moderate positive impact on financial performance. TQM practices and non-financial performance Although financial performance is generally accepted as the ultimate aim of business organizations, in the case of SMEs, non-financial performance indicators are also equally important in implementing TQM principles. TQM practices may not only affect financial performance directly (Kaynak, 2003), but also in some indirect ways such as increasing innovation (Singh and Smith, 2004), changing organizational culture (Irani et al., 2004), market competitiveness (Chong and Rundus, 2004), overall organizational performance (Powel, 1995), market share and market share growth (Kaynak, 2003); employee morale (Rahman and Bullock, 2005), productivity (Rahman and Bullock, 2005; Kaynak, 2003; Rahman, 2001). Prajogo and Sohal (2001) report two main arguments on the relationship between TQM and innovation where the first argument suggests that TQM be positively related to increasing innovation capacity of TQM practicing firms. The second argument, however, focuses on the negative

TQM and organizational performance 833

JMTM 17,6

834

relationship between TQM implementation and innovative performance of firms. The logic behind this argument is that customer focus and its principles may trap organizations into captive markets where they focus only on existing customers, which may result in ignoring the search for innovation and novel solutions (Prajogo and Sohal, 2006). Samson and Terziovski (1999) found support for the relationship between some non-financial measures (i.e. export growth, market share growth, innovation growth, cost of quality, etc.) and implementation of TQM practices. In the case of SMEs, however, the evidence is sketchy. Choi and Eboch (1998) argue that the strength of positive relationship between plant performance, as influenced by TQM practices, and customer satisfaction, is still far from being conclusive. Samson and Terziovski (1999) noted negative relationship for smaller size firms in their survey, whereas Lee (2004) reports that Chinese SMEs perceive positive relationship between TQM practices and non-financial performance measures (i.e. production performance, cost improvement and sales improvement). While there is no detailed analysis of the relationship between TQM practices and non-financial performance for SMEs in the prior literature (Ahire and Golhar, 1996), we rely on the argument of the first group of researchers and assert the following hypothesis: H2. TQM practices have a strong positive impact on non-financial performance. Relationship between financial and non-financial performance The relationship between financial and non-financial measures of organizational performance has long been discussed in organization and strategy literature. Hackman and Wageman (1995) provide an insightful account of conceptual and practical issues in researching TQM implementation and change. York and Miree (2004) argue that non-financial performance such as improved quality, innovativeness and increased market share should actually reduce costs, and thus have a positive effect on measures of financial performance. Increased quality helps SMEs to retain current customers and create greater customer loyalty, which in return may increase market share and financial performance (Rust et al., 1994). Although studies of SME performance and TQM relations do not examine non-financial performance measures directly, evidence from larger organisations supports the argument that operational performance indicators are related to financial performance dimensions (Fuentes-Fuentes et al., 2004). Some other studies also demonstrate positive relationship between operational performance dimensions such as product quality (Larson and Sinha, 1995), innovation and R&D (Prajogo and Sohal, 2001; Singh and Smith, 2004) employee performance (Fuentes-Fuentes et al., 2004). Given the availability of conclusive evidence supporting a relationship between the non-financial performance dimensions and financial performance, we hypothesize that: H3. Non-financial performance has a strong impact on financial performance. Mediating effect of non-financial performance Earlier studies of TQM implementation and financial performance treated TQM elements as independent variables and tried to establish relationship between them. In the case of SMEs, non-financial performance dimensions are equally significant and may intermediate between TQM practices and financial performance (York and Miree, 2004; Rahman and Bullock, 2005) hence indicating effectiveness of implementation

(Prajogo and Sohal, 2006). Hackman and Wageman (1995) point out long-term versus short-term performance issues in TQM implementing organisations. This argument also can be extended to cover TQM implementation in SMEs, where there may be time lag between implementation and financial performance. Based on these arguments and supporting evidence we expect that:

TQM and organizational performance

H4. TQM practices have a strong positive impact on financial performance with a mediating effect of non-financial performance.

835

These theoretical discussions and proposed hypothesized relationships are delineated in the following research model, as shown in Figures 1 and 2. δ1 Q1 δ2 Q 2 λ11 x

δ3 Q3

λ21 x λ31

δ4 Q4

λ41 x λ51 x

δ5 Q5

λ61

λ12 y

x

Total Quality Management

Financial Performance

Y11

ξ1

η1

z1

x

λ22 y

F1

ε1

F2

ε2

F3

ε3

F4

ε4

F1

ε6

F2

ε7

F3

ε8

F4

ε9

λ32 y λ42 y

λ71 x

δ6 Q 6 δ7 Q 7

δ1

Q1

δ2

Q2 λ11x

δ3

Q3 Q4

λ41x λ51x

δ5

Q5

λ61x λ71x

δ6

Q6

δ7

Q7

ε2 NF2

λ21x λ31x

δ4

ε1 NF1

ε3 NF3

ε4

ε5

NF4

NF5

λ11y λ 21y λ 31y λ 41y λ 51y

Total Quality Management

ξ1

Y11

Non-Financial Performance

η1

ζ1

Y21

λ62 y

Financial Performance

η2

ζ2

λ72 y λ82 y λ92 y

Figure 1. The structural relationship between TQM practices and financial performance

Figure 2. The structural relationship between TQM practices and financial performance with a mediation of non-financial performance

JMTM 17,6

836

Research methodology Survey instrument The survey instrument used in this study was largely derived from the work of Saraph et al. (1989) with the purpose of identifying critical factors of TQM in a business unit. In the present questionnaire, the initial set of 30 items was reduced to 20. The basic justification for this lies in the researchers’ impression (derived from the pilot study) that the SMEs are in the “awakening” stage described by Crosby (1996). Our interviews corroborated that management “recognized that quality management may be of value but was not willing to provide money or allocate time to make it all happen, teams were set up to attack major problems instead of soliciting long range solutions” and that company quality posture could be summarized as “is it absolutely necessary to always have problems with quality?” These signified a very close alignment with the “awakening” stage of Crosby’s stages of maturity. The original version of the questionnaire was in English. This questionnaire was translated into the local language (Turkish). The local version was back translated until a panel of experts agreed that the two versions were comparable. Each item was rated on a five-point Likert scale, ranging from “very low” to “very high”. The questionnaire was pre-tested several times to ensure that the wording, format, and sequencing of questions were appropriate. The extent of TQM implementation and the level of organizational performance on both financial and non-financial performance measures were determined using judgmental measures based on managers’ perceptions of how the organization was performing on each constituent item. As the percentage of missing data was calculated to be relatively small, occasional missing data on variables was handled by replacing them with the mean value. The measures of TQM practices and organizational performance are provided in Appendices 1 and 2. The sample There is no consensus on the definition of SME, as variations exist between countries, sectors and even different governmental agencies within the same country (Yusof and Aspinwall, 2000). In line with small business research, this study adopted the number of employees as the base for the definition of SME. An SME is identified as one that employs fewer than 100 persons. The minimum of at least ten employees was also chosen in order to exclude very minor firms that would not be suitable for the purposes of this study. This range is consistent with the definition of an SME adopted by both the Turkish State Institute of Statistics (SIS) and Turkish Small Business Administration and also by a number of European countries such as Norway and Northern Ireland (Sun and Cheng, 2002; McAdam and McKeown, 1999). Data for this study were collected using a self-administered questionnaire that was distributed to 500 SMEs in textile industry in the city of Istanbul in Turkey selected randomly from the database of Turkish Small Business Administration (KOSGEB). As of 2005, the KOSGEB database includes a total of 12,270 SMEs in Istanbul, which accounts for nearly 28 per cent of all SMEs registered throughout Turkey. The sampling frame consists of 2,482 SMEs operating in the textile industry in Istanbul. The study focused on the textile industry including textile mill products and apparel (SIC codes 22 and 23), since it has been a leader in implementing progressive quality management practices in Turkey. The textile industry has also been the engine of economic growth and generates the largest volume of export revenues. Although one

could argue that a focus on a single industry may make the results less generalizable, we ensured a high level of internal validity. Furthermore, within the textile industry there exist several different manufacturing environments and product types making the sample much more diverse than what could be expected for a homogenous sample. It was requested that the questionnaire be completed by a senior officer/executive in charge of quality management. The responses indicated that a majority of the respondents completing the questionnaire were in fact members of the top management. Of the 500 questionnaires posted, a total of 163 questionnaires were returned after one follow-up. About 22 questionnaires were eliminated due to largely missing values. The overall response rate was thus 28 per cent (141/500), which was considered satisfactory for subsequent analysis. A comparison of the annual sales volume, number of employees and sub-industry variation revealed no significant differences between the responding and non-responding firms ( p . 0.1). Thus, the responses adequately represented the total sample group. Results The data analysis is conducted at three steps: (1) Performing an exploratory factor analysis (EFA) with varimax rotation to determine the underlying dimensions of TQM. (2) Testing of the measurement models for TQM construct using confirmatory factor analysis (CFA) as well as the TQM context in order to determine if the extracted dimensions in step 1 offered a good fit to the data. (3) Measuring the impact of critical factors of TQM on business financial performance. These steps are discussed in the following subsections. Exploratory factor analysis EFA with varimax rotation was performed on the TQM criteria in order to extract the dimensions underlying the construct. The EFA of the 20 variables yielded seven factors explaining 78.6 per cent of the total variance. Based on the items loading on each factor, these factors were labelled as “quality data and reporting” (factor 1), “role of top management” (factor 2), “employee relations” (factor 3), “supplier quality management” (factor 4), “training” (factor 5), “quality policy of top management” (factor 6) and “process management” (factor 7). These items are shown in Table I. The Cronbach’s alpha measures of reliability for the seven factors vary between 0.75 and 0.86, which are well above the traditionally acceptable value of 0.70, as shown in Table I. Confirmatory factor analysis This stage is also known as testing the measurement model where TQM was tested using the first order confirmatory factor model to assess construct validity. The results consistently supported the factor structure for TQM that is shown by the EFA. Table II summarizes the measurement models for TQM practices and shows the standardized regression weight for each variable. The factor level analysis is a separate analysis for each factor including only the indicators for that factor (i.e. variables loading on that factor).

TQM and organizational performance 837

JMTM 17,6

Symbol Q2 Q1

838

Q4 Q3 Q5 Q6 Q7 Q8 Q9 Q10 Q12 Q13 Q11 Q14 Q15 Q16 Q17 Q18

Table I. Exploratory factor analysis of TQM practices

Q19 Q20

Variables

1

Extent to which quality data are available to managers and supervisors Extent to which quality data are used as tools to manage quality Scope of the quality data includes process/service performance Extent to which quality data are used to evaluate supervisors and managerial performance Acceptance of responsibility for quality by major department heads Extent to which top management supports a long term quality improvement process Extent to which top management has objectives for quality performance Amount of feedback provided to the employees on their quality performance Degree of participation in quality decisions by employees Extent to which employees are recognized for superior quality performance Clarity of specifications provided to suppliers Evaluation of performance of suppliers Extent to which longer term relationships are offered to suppliers Training in advanced techniques Training in statistical techniques Specific work-skill training Importance attached to quality by top management in relation to cost/revenue objectives Degree to which top management considers quality improvement as a way to increase profits Importance of inspections, review or checking of work Amount of inspections, review or checking of work

0.80

2

3

Factors 4 5

6

7

0.75 0.74 0.71 0.76 0.76 0.73 0.76 0.72 0.51 0.81 0.80 0.74 0.82 0.82 0.67 0.87 0.75 0.80 0.67

The standardized regression weights for all variables that are shown in Table II are significant at the 0.05 level. The CFA showed a good fit. The x2 statistic was 177.3 (degrees of freedom ¼ 149, p , 0.05), with the x 2/df ratio having a value of 1.19 that is less than 2.0 (it should be between 0 and 3 with lower values indicating a better fit). The goodness of fit index (GFI) was 0.89 and adjusted goodness of fit (AGFI) index was 0.85. These scores are very close to 1.0 (a value of 1.0 indicates perfect fit). The comparative fit index (CFI) was 0.98, Tucker-Lewis coefficient (TLI) was 0.92. All indices are close to a value of 1.0 in CFA indicating that the measurement models provide good support for the factor structure determined through the EFA. The model parameters were estimated using the method of maximum likelihood. The average of item scores for each factor in TQM construct was used as measures in the path model.

Symbol

Description

Quality data and reporting Q2 Extent to which quality data are available to managers and supervisors Q1 Extent to which quality data are used as tools to manage quality Q4 Scope of the quality data includes process/service performance Q3 Extent to which quality data are used to evaluate supervisors and managerial performance Role of top management Q5 Acceptance of responsibility for quality by major department heads Q6 Extent to which top management supports a long term quality improvement process Q7 Extent to which top management has objectives for quality performance Employee relations Q8 Amount of feedback provided to the employees on their quality performance Q9 Degree of participation in quality decisions by employees Q10 Extent to which employees are recognized for superior quality performance Supplier quality management Q12 Clarity of specifications provided to suppliers Q13 Evaluation of performance of suppliers Q11 Extent to which longer term relationships are offered to suppliers Training Q14 Training in advanced techniques Q15 Training in statistical techniques Q16 Specific work-skill training Quality policy Q17 Importance attached to quality by top management in relation to cost/revenue objectives Q18 Degree to which top management considers quality improvement as a way to increase profits Process management Q19 Importance of inspections, review or checking of work Q20 Amount of inspections, review or checking of work

Regression weight

t-value

0.80

8.72

0.71 0.70 0.90

7.78 – 9.59

0.71

8.21

0.78

9.04

0.79

–

0.70

7.42

0.66 0.75

7.09 –

0.77 0.87 0.72

8.31 8.96 –

0.81 0.88 0.67

8.11 8.45 –

0.73

8.04

0.92

–

0.70 0.91

– 8.33

Note: – Fixed for estimation

Unidimensionality tests of constructs in the path model The validity and reliability of three constructs were assessed by checking unidimensionality of each construct using three tools: Principal component analysis, Cronbach’s a and Dillon-Goldstein’s r. As shown in Table III, all of the Cronbach’s a values met the threshold value of 0.70. In line with the principal component analysis, since the first eigenvalue score of the correlation matrix of the manifest variables of each construct is larger than one and the second one is smaller than one then the construct was considered as unidimensional. Similarly, Dillon-Goldstein’s r analysis provides r values above 0.70 for each construct supporting unidimensionality.

TQM and organizational performance 839

Table II. Confirmatory factor analysis of TQM practices

JMTM 17,6

840

Path model The final step in the analysis was to test the path model as specified in Figures 1 and 2. The hypothesized structural equation model was tested using LISREL software package (Jo¨reskog and So¨rbom, 1993). Model fit determines the degree to which the structural equation model fits the sample data. Model fit criteria commonly used are chi-square (x 2), GFI, AGFI and root mean square residual (RMS) (Schumacker and Lomax, 1996). The goodness-of-fit indices for the first model (Figure 1) (Satorra-Bentler x 2 ¼ 54.89 with df ¼ 41; GFI ¼ 0.93; AGFI ¼ 0.89; NFI ¼ 0.92; CFI ¼ 0.98) and the second structural model (Figure 2) (Satorra-Bentler x 2 ¼ 114.29 with df ¼ 95; GFI ¼ 0.91; AGFI ¼ 0.87; NFI ¼ 0.90; CFI ¼ 0.98) are well within the generally accepted limits. Hence, both models were accepted to fit the data. Hypothesis test results H1: Relationship between TQM and financial performance. Figure 3 shows the model that is related to the first hypothesis and is provided in the following equations.

h1 ¼ g11 j1 þ z1 Fin ¼ 0:24

Block Table III. Unidimensionality of constructs

Number of indicators

Cronbach a

Dillon-goldstein’s r

First eigenvalue

Second eigenvalue

7 5 4

0.860 0.856 0.836

0.894 0.899 0.894

2.787 2.701 2.126

0.635 0.420 0.498

TQM NONFIN FIN

0.37

TQM1

0.37

TQM2

TQM þ z1

F1

0.67 0.34

TQM3

0.49

0.80

0.65 0.36

TQM4

0.54 0.65

0.49

Figure 3. Inner and outer regression weights for the structural relationship between TQM practices and financial performance

TQM5

0.44 0.59

0.54

TQM6

0.27

TQM7

Total Quality Management ξ1

0.24

Financial Performance η1

0.23

F2

0.10

F3

0.35

F4

0.53

0.76 0.69 0.37

The first model has one endogenous variable (dependent variable), labelled as financial performance and one exogenous variable (independent variable), which is labelled as TQM. TQM explains almost 5.6 percent of the variation in financial performance. The hypothesized relationship was tested using the associated t-statistics. The standardized regression weight for H1 was found to be 0.24 ( p , 0.05). Although t value is significant, TQM explains only a small percentage of the variation in financial performance. Therefore, a good deal of support has been provided to H1 that TQM practices have a moderate positive impact on financial performance. H2: TQM practices and non-financial performance. The relationship between TQM practices and non-financial performance that is shown in Figure 4 is presented in the following equation.

TQM and organizational performance 841

Nonfin ¼ 0:67 TQM þ z1 The second model has one endogenous variable (dependent variable), which is named Non-Financial Performance, and one exogenous variable (independent variables), which is labelled as TQM. Nearly 45 percent of the variation in financial performance was explained by TQM construct. With respect to the second hypothesis (H2), it was found that TQM had a significant positive impact on non-financial performance. H3: relationship between financial and non-financial performance. The endogenous variable of third model is Financial Performance and exogenous variable is non-financial performance. Non-financial performance explains almost 40 percent of the variation in financial performance. The standardized regression weight was found to be 0.63 which is significant at 0.01 level indicating a strong support for H3 that non-financial performance has a strong impact on financial performance, as shown in Figure 4. H4: mediating effect of non-financial performance. A good deal of support has been found for H4 that TQM practices have a strong positive impact on financial performance with a mediating effect of non-financial performance as shown in equation (3) where it is obtained by substituting equation (1) in equation (2).

0.41

TQM1

0.37

TQM2

0.64 0.30

TQM3

0.41

0.36

0.39

0.38

0.34

NF1

NF2

NF3

NF4

NF5

0.64 0.70

0.57 0.72

0.50

0.80

0.75

0.67 0.32

TQM4

0.57 0.69

0.43

TQM5

0.54

TQM6

0.32

TQM7

0.44 0.54

Total Quality Management ξ1

0.67

Non-Financial Performance η1

0.63

Non-Financial Performance η2

F1

0.32

F2

0.088

F3

0.28

F4

0.50

0.80 0.75 0.42

Figure 4. Inner and outer regression weights for the structural relationship between TQM practices and financial performance with a mediation of non-financial performance

JMTM 17,6

842

Nonfin ¼ 0:67 TQM þ z1

ð1Þ

Fin ¼ 0:63 Nonfin þ z2

ð2Þ

Fin ¼ 0:422 TQM þ z1

ð3Þ

Conclusion and managerial implications The findings of this study reveal that indirect effect of the critical factors of TQM on financial performance mediated by non-financial performance has more influence than direct effect of the critical factors of TQM on financial performance. These findings indicate that financial performance measures such as revenue, net profits, return on assets, and profit to revenue ratio are partially explained by the implementation of TQM practices. On the other hand, TQM practices provide a better explanation on financial performance through non-financial performance criteria such as market development, market orientation and investment in R&D. In a similar vein, nearly forty per cent of variation in financial performance is explained indirectly by quality practices. The most important quality practices were found to be training, employee relations, and quality data and reporting. Hence, companies should be suggested to develop formal reward and recognition systems to encourage employee involvement and participation, support teamwork and provide feedback to the employees. There are many purposes for gathering data in quality management. First, data can be collected to understand current processes. Data also provide inspection, various test results and verification records. They are used to analyse the process using various types of statistical process control tools such as control charts, Pareto charts, cause and effect diagrams, check sheet, histograms, and scatter diagram among others. These traditional quality tools are very useful in monitoring and measuring progress and performance. Data analysis and interpretation may also help SMEs to develop a learning environment, which may enhance innovation and a better enterprise culture. Management by facts requires management decisions to be based on relevant data and reports. Based on the study’s findings, the least important factor was found to be the role of top management. In fact, success of TQM applications hinges on strong leadership that must be initiated by the top management. Quality improvement plans proposed by several gurus strongly emphasize the commitment of top management. The top management of the organization is directly responsible for determining an appropriate organization culture, vision, and quality policy. Top managers should also determine objectives, and develop specific and measurable goals to satisfy customer expectations and improve their organizations’ performance. In order to enhance net profit and revenue as well as to reduce cost of quality, managers must convey their priorities and expectations to their employees. In SMEs, owner of the company usually does not delegate adequate power and responsibility to top managers of the company. The basic justification for this lies in the researchers’ impression (derived from the pilot study) that the SMEs are largely in the “awakening” stage as described by Crosby (1996). This appears to be a major weakness in TQM implementing SMEs, which may not be unique only to Turkey. While most SMEs in Turkey do not have an established quality department, many of them have invested substantial resources in adapting and implementing TQM

programs to improve their performance. It is generally accepted that several SMEs did not achieve any improvement and some only a little. Specifically, due to the presence of a multitude of barriers, many organizations utilize only a partial implementation of TQM, and hence are unable to achieve continuous and systematic improvement. If TQM plan is implemented properly, it produces a variety of benefits such as understanding customers’ needs, improved customer satisfaction, improved internal communication, better problem solving and fewer errors. The success of a TQM program increases when its implementation is extended to the entire company. This enables the reformation of the corporate culture and the permeation of the new business philosophy into every facet of organization. The philosophy of doing things right must be implemented with enthusiasm and commitment throughout the organization -from top to bottom and the little steps forward the so-called “Kaizen” by the Japanese- must be viewed as “a race without a finish”. Consequently, effective implementation of TQM is a valuable asset in a company’s resource portfolio, one that can produce important competitive capabilities and be a source of competitive advantage. Performance is a multifaceted concept and this study tried to capture performance dimensions from both financial non-financial perspectives. Number of other factors (both internal and external to organisation) may also mediate TQM implementation and performance relationship. Although this study establishes relationship between TQM implementation and performance dimensions, other factors such as size, culture, absorptive and innovative capacities and market orientation of sample firms may also have some impact on organisational performance. Market orientation and innovation (may be embedded in an organisational culture) seem to be highly relevant to TQM implementation and performance which merit further research on SMEs. It should also be acknowledged that the study is subject to some methodological limitations. First, it would be highly suggested that the size and nature of the sample must be enhanced to ensure variability and control for possible extraneous variation. While the sample is restricted to only a single region and a single industry, it would be strongly recommended that data should be gathered from various parts of Turkey including both various manufacturing and service industries. Since, the data in this study were collected from top managers of organizations on the basis of their subjective evaluations, objective performance indicators should also be employed in the analysis. Finally, neural network model could be utilized in the future studies to gain additional insights in exploring the relationship between TQM and organizational performance. References Ahire, S.L. and Golhar, D.Y. (1996), “Quality management in large vs small firms”, Journal of Small Business Management, Vol. 34 No. 2, pp. 1-11. Ahire, S.L., Golhar, D.Y. and Waller, M.A. (1996), “Development and validation of TQM implementation constructs”, Decision Sciences, Vol. 27 No. 1, pp. 23-56. Anderson, J.C., Rungtusanatham, M. and Schroeder, R.G. (1994), “A theory of quality management underlying the Deming management method”, Academy of Management Review, Vol. 19 No. 3, pp. 472-509. Benson, G.P., Saraph, J.V. and Schroeder, R.G. (1991), “The effects of organizational context on quality management: an empirical investigation”, Management Science, Vol. 37 No. 9, pp. 1107-24.

TQM and organizational performance 843

JMTM 17,6

844

Black, S.E. and Porter, L.J. (1996), “Identification of the critical factors of TQM”, Decision Sciences, Vol. 27 No. 1, pp. 1-21. Cagliano, R., Blackmon, K. and Voss, C. (2001), “Small firms under MICROSCOPE: international differences in production/operations management practices and performance”, Integrated Manufacturing Systems, Vol. 12 No. 7, pp. 469-82. Choi, T.Y. and Eboch, K. (1998), “The TQM paradox: relations among TQM practices, plant performance and customer satisfaction”, Journal of Operations Management, Vol. 17, pp. 59-75. Chong, V.K. and Rundus, M.J. (2004), “Total quality management, market competition and organizational performance”, British Accounting Review, Vol. 36, pp. 155-72. Crosby, P.B. (1996), Quality is Free, McGraw-Hill, New York, NY. Dangayach, G.S. and Desmukh, S.G. (2005), “Advanced manufacturing technology implementation: evidence from Indian small and medium enterprises”, Journal of Manufacturing Technology Management, Vol. 16 No. 5, pp. 483-96. Deming, W.E. (1986), Out of the Crisis, MIT Centre for Advanced Engineering Study, Cambridge, MA. Douglas, T.J. and Judge, W.Q. Jr (2001), “Total quality management implementation and competitive advantage: the role of structural control and exploration”, Academy of Management Journal, Vol. 44 No. 1, pp. 158-69. Easton, G.S. and Jarrel, S.L. (1998), “The effects of total quality management on corporate performance: an empirical investigation”, Journal of Business, Vol. 71 No. 2, pp. 253-307. Flynn, B.B. and Saladin, B. (2001), “Further evidence on the validity of the theoretical models underlying the Baldrige criteria”, Journal of Operations Management, Vol. 19, pp. 617-52. Flyyn, B.B., Schroder, R.G. and Sakakibara, S. (1994), “A framework for quality management research and an associated measurement instrument”, Journal of Operations Management, Vol. 11, pp. 339-66. Flyyn, B.B., Schroeder, R.G. and Sakakibara, S. (1995), “The impact of quality management practices on performance and competitive advantage”, Decision Sciences, Vol. 26 No. 5, pp. 659-91. Fuentes-Fuentes, M.M., Albacate-Saez, C.A. and Llorens-Montes, F.J. (2004), “The impact of environmental characteristics on TQM principles and performance”, Omega, Vol. 32 No. 6, pp. 425-42. Fynes, B. and Voss, C. (2001), “A path analytic model of quality practices, quality performance and business performance”, Production and Operations Management, Vol. 10 No. 4, pp. 494-513. Gadenne, D. and Sharma, B. (2002), “An inter-industry comparison of quality management practices and performance”, Managing Service Quality, Vol. 12 No. 6, pp. 394-404. Garvin, D.A. (1987), “Competing on the eight dimensions of quality”, Harvard Business Review, pp. 101-9, November/December. Hackman, R.J. and Wageman, R. (1995), “Total quality management: empirical, conceptual and practical issues”, Administrative Science Quarterly, Vol. 40 No. 2, pp. 309-42. Hendricks, K.B. and Singhal, V.R. (1996), “Quality awards and the market value of the firm: an empirical investigation”, Management Science, Vol. 42 No. 3, pp. 415-36. Hendricks, K.B. and Singhal, V.R. (2001), “Firm characteristics, total quality management, and financial performance”, Journal of Operations Management, Vol. 19, pp. 269-85.

Irani, Z., Beskese, A. and Love, P.E.D. (2004), “Total quality management and corporate culture: constructs of organizational excellence”, Technovation, Vol. 24, pp. 643-50. Jo¨reskog, K.G. and So¨rbom, D. (1993), LISREL8: Structural Equation Modelling with the SIMPLIS Command Language, Lawrence Erlbaum Association, Hillsdale, NJ. Kanji, G.K. and Wallace, W. (2000), “Business excellence through customer satisfaction”, Total Quality Management, Vol. 11 No. 7, pp. 979-98. Kaplan, R.S. and Norton, D.P. (1996), The Balanced Scorecard: Translating Strategy into Action, Harvard Business School Press, Boston, MA. Kaynak, H. (2003), “The relationship between total quality management practices and their effects on firm performance”, Journal of Operations Management, Vol. 21, pp. 405-35. Larson, P.D. and Sinha, A. (1995), “The total quality management impact: a study of quality managers’ perceptions”, Quality Management Journal, Vol. 2 No. 3, pp. 53-66. Lee, C.Y. (1998), “Quality management by small manufacturers in Korea: an exploratory study”, Journal of Small Business Management, Vol. 36 No. 4, pp. 61-7. Lee, C.Y. (2004), “Perception and development of total quality management in small manufacturers: an exploratory study in China”, Journal of Small Business Management, Vol. 42 No. 1, pp. 102-15. McAdam, R. and McKeown, M. (1999), “Life after ISO 9000: an analysis of the impact of ISO 9000 and total quality management on small businesses in Northern Ireland”, Total Quality Management, Vol. 10 No. 2, pp. 229-41. Madu, C.N., Kuei, C.H. and Jacob, R.A. (1996), “An empirical assessment of the influence of quality dimensions on organizational performance”, International Journal of Production Research, Vol. 34 No. 7, pp. 1943-62. Mitra, A. (1987), Fundamentals of Quality Control and Improvement, Prentice-Hall, Englewood Cliffs, NJ. Montes, F.J.L.M., Jover, A.V. and Fernandez, L.M.M. (2003), “Factors affecting the relationship between total quality management and organizational performance”, International Journal of Quality & Reliability Management, Vol. 20 No. 2, pp. 189-209. Powel, T.C. (1995), “Total quality management as competitive advantage: a review and empirical study”, Strategic Management Journal, Vol. 16, pp. 15-37. Prajogo, D.I. and Sohal, A.S. (2001), “TQM and innovation: a literature review and research framework”, Technovation, Vol. 21, pp. 539-58. Prajogo, D.I. and Sohal, A.S. (2006), “The relationship between organization strategy, total quality management (TQM) and organization performance-the mediating role of TQM”, European Journal of Operational Research, Vol. 168 No. 1, pp. 35-50. Rahman, S. (2001), “A comparative study of TQM practice and organisational performance of SMEs with and without ISO 9000 certification”, International Journal of Quality and Reliability Journal, Vol. 18 No. 1, pp. 35-49. Rahman, S. and Bullock, P. (2005), “Soft TQM, hard TQM and organizational performance relationships: an empirical investigation”, Omega, Vol. 33, pp. 73-83. Raymond, L. (2005), “Operations management and advanced manufacturing technologies in SMEs”, Journal of Manufacturing Technology Management, Vol. 16 No. 8, pp. 936-55. Rust, R.T., Zahorik, A.J. and Keiningham, T.L. (1994), Return on Quality: Measuring the Financial Impact of Your Company’s Request for Quality, Probus Publishing Company, Chicago, IL.

TQM and organizational performance 845

JMTM 17,6

846

Samson, D. and Terziovski, M. (1999), “The relationship between total quality management practices and operational performance”, Journal of Operations Management, Vol. 17, pp. 393-409. Saraph, J.V., Benson, G.P. and Schroder, R.G. (1989), “An instrument for measuring the critical factors of quality management”, Decision Sciences, Vol. 20, pp. 810-29. Schmidt, W. and Finnigan, J. (1992), The Race without a Finish Line: America’s Quest for Total Quality, Jossey-Bass, San Francisco, CA. Schumacker, R.E. and Lomax, R.G. (1996), A Beginner’s Guide to Structural Equation Modelling, Lawrence Erlbaum Associates, Mahwah, NJ. Singh, P.J. and Smith, A.J.R. (2004), “Relationship between TQM and innovation: an empirical study”, Journal of Manufacturing Technology Management, Vol. 15 No. 5, pp. 394-401. Strubering, L. and Klaus, L.N. (1997), “Small business thinking big”, Quality Progress, Vol. 2, pp. 23-7. Sun, H. and Cheng, T.K. (2002), “Comparing reasons, practices and effects of ISO 9000 certification and TQM implementation in Norwegian SMEs and large firms”, International Small Business Journal, Vol. 20 No. 4, pp. 421-40. Taymaz, E. (1997), Small and Medium-Sized Industry in Turkey, SIS, Ankara. UNCTAD (1993), Small- and Medium-sized Transnational Corporations, United Nations, New York, NY. Wilson, D.D. and Collier, D.A. (2000), “An empirical investigation of the Malcolm Baldrige National Quality Award causal model”, Decision Sciences, Vol. 31 No. 2, pp. 361-90. York, K.M. and Miree, C.E. (2004), “Causation or covariation: an empirical re-examination of the link between TQM and financial performance”, Journal of Operations Management, Vol. 22, pp. 291-311. Yusof, S.M. and Aspinwall, E. (2000), “Total quality management implementation frameworks: comparison and review”, Total Quality Management, Vol. 11 No. 3, pp. 281-94. Zairi, M. and Ahmed, P.K. (1999), “Benchmarking maturity as we approach the millennium”, Total Quality Management, Vol. 10 Nos 4/5, pp. 810-6. Appendix 1. TQM practices (1) Extent to which quality data are used as tools to manage quality (2) Extent to which quality data are available to managers and supervisors. (3) Extent to which quality data are used to evaluate supervisor and managerial performance. (4) Scope of the quality data includes process/service performance. (5) Acceptance of responsibility for quality by major department heads. (6) Extent to which top management supports long-term quality improvement process. (7) Extent to which the top management has objectives for quality performance. (8) Amount of feedback provided to the employees on their quality performance. (9) Degree of participation in quality decisions by hourly/non-supervisory employees. (10) Extent to which employees are recognized for superior quality performance. (11) Extent to which longer term relationships are offered to suppliers. (12) Clarity of specifications provided to suppliers. (13) Assessment of performance of suppliers.

(14) (15) (16) (17) (18)

Training in advanced technique. Training in statistical technique. Specific work skill training. Importance attached to quality by top management in relation to cost/revenue objectives. Degree to which top management considers quality improvement as a way to increase profits. (19) Importance of inspection, review or checking of work. (20) Amount of inspection, review or checking of work. Appendix 2. Organizational performance (1) Financial performance. . Revenue growth over the last three years. . Net profits. . Profit to revenue ratio. . Return on assets. (2) Non-financial performance. . Investments in R&D aimed at new innovations. . Capacity to develop a unique competitive profile. . New product/service development. . Market development; and . Market orientation. Corresponding author Mehmet Demirbag can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

TQM and organizational performance 847

The current issue and full text archive of this journal is available at www.emeraldinsight.com/1741-038X.htm

JMTM 17,6

Transforming the supply chain Evangelia D. Fassoula Neoset S.A., Athens, Greece University of Pireaus, Athens, Greece

848 Received October 2005 Revised January 2006 Accepted March 2006

Abstract Purpose – Today’s business environment emerges the need for organizations to continuously transform themselves, in order to maintain and reinforce their ability to compete successfully. The purpose of this paper is to present an analysis of the transformation process of supply chain in order to provide a modular structured management tool for planning, implementing and measure the effectiveness of supply chain transformation process (SCTP), in relation to overall organizational performance and business strategy. Design/methodology/approach – The focal concepts upon which the overall approach is based are the strategic planning process, the human resources management techniques, the link between organizational structure and culture, the marketing integration and the importance of quality evaluation. Findings – The success of such efforts depends on a number of correlated parameters, among which, those of top management commitment and behavioural issues of human resources are the most critical. The tool proposed in this paper for the transformation of the supply chain, provides a flexible sequence of project phases, linked together in a loop and continuously assessed against specific evaluation criteria, to assure the project quality and the continual transformation dynamics. Practical implications – The pilot implementation of the tool to a big furniture organization led to the development of an integrated supply chain function, which would integrate planning, purchasing, production, warehousing distribution and transportation processes. The customer satisfaction rate had an increase of 25 percent, six months after the fulfilment of the transformation and three key suppliers reported a significant decrease of the duration of the overall purchasing process. Originality/value – The paper presents a modular tool to facilitate the management of a SCTP as a project, taking under consideration the critical sensitivity factors of a transformation process. Keywords Supply chain management, Organizational processes, Competitive strategy Paper type Conceptual paper

Journal of Manufacturing Technology Management Vol. 17 No. 6, 2006 pp. 848-860 q Emerald Group Publishing Limited 1741-038X DOI 10.1108/17410380610678837

Introduction During the last two decades, we can see a great number of organizations making efforts to change some or all of their core business processes, by using the total quality management approach or the business process reengineering techniques, or simply by re-examining their organizational structure, the way business processes are planned, performed and measured and also the way human resources are managed. What is really different today, as we are moving to the middle of the first decade of the twenty-first century is that the global business environment and social and market structures are rapidly and, some times, unpredictably changing. Customer needs and expectations are becoming more and more unstable and difficult to identify and, at the same time, customers have many choices, frequently similar in terms of quality and price. Differentiation and/or cost reduction strategies are embodied by organizations in order to create competitive advantage. The overall context is indicating the need for all kinds of organizations not just to rethink the way they are structured and operate and determine a changed model; they must re-invent themselves in a way that can allow

them to continuously adapt to different market requirements and compete successfully. Organizations must get used to the fact that the new competition will force them to “burn themselves down” and rebuild every few years, Martin (1993) mentions. The supply chain process is a core business process of major importance for the realization of business strategy. It determines numerous key performance indicators of an organization and has a major impact in its profitability and competitiveness. Therefore, supply chain can be considered as maybe the most suitable operational framework for a transformation process to be based on. The objective of this paper is to contribute to the efforts of developing and maintaining competitive advantage, by analyzing the way transformation process is influencing and linked with supply chain and providing a management tool for planning, implementing and measuring the effectiveness of supply chain transformation process (SCTP), in relation to overall organizational performance. Overall organizational performance is meant to reflect the satisfaction rate of all interested parties (customers, employees, stakeholders, suppliers and social partners) and the reliability of operations, as well as to reflect qualities like openness, flexibility, adaptation and quick market response. The analysis of the transformation process focuses on the following distinct supply chain processes, which are interrelated: order processing, procurement, inventory management (raw materials – in process – finished goods), warehousing, transportation, distribution, customer service and reverse logistics. Such a transformation cannot be studied separately form the strategic management framework of the organization and cannot be managed without being considered as a project. The management tool has a modular structure, according to the distinct supply chain processes and its implementation is based on a sequence of project phases. The tool is complemented with detailed practical guidance for each project phase and criteria of successful performance for carrying out an on-going progress evaluation procedure. The critical sensitivity factors of a transformation process (strategic planning, human resources management, marketing integration and organizational culture) are the conceptual platform of the tool. Background The strategic management framework Change efforts cannot be considered separate from the strategic management framework of an organization. Unless that, problems of integration may occur as the change process will evolve. Additionally, without a solid change strategy being reflected to the tactical and operational level, the change project may be fragmented between contradictory business priorities. The relationship between organizational strategy and structure has been analyzed a lot as it can be seen by reviewing the relevant literature. According to Chandler (1962), in the 1950s, when the demand for manufactured goods often exceeded their supply and organizations were “pulled” along by consumers anxious to raise their standard of living, the structure of an organization considered to follow its strategy. As decades passed, markets have changed, with the result that supply usually exceeds demand and a much more complex business environment has been created. Based on Drucker (1988), strategy is no longer seen as a logical detached process and is increasingly seen as something that emerges from the organization’s efforts to stay competitive, so the

Transforming the supply chain

849

JMTM 17,6

850

idea that a structure will in some way contribute to the creation of a strategy has grown in importance. In all cases, strategy and structure should be closely linked and, this becomes much more significant since the requirement of flexibility is now and imperative for all organizations, as well as their quick response to continuously changing customer needs with new products and/or services. Figure 1 shows the components of the strategic management framework, indicated by Hax et al. (1996). The figure communicates the interrelationship that links strategy, structure, processes, performance and culture. It is important to underline three issues: (1) The impact of culture (which means the human parameter and related behavioural issues) on strategy and vice versa. (2) The fact that the information and communication process serve as coordinating media in the organization structure. (3) The business processes capture the interdependence of the tasks across the organizational units defined by structure. From the brief analysis presented, it can be resulted that the organization structure is part of a dynamic integrated system of parameters, closely linked with strategy and the human parameter and related to behavioural issues, and any change effort has to carefully manage the balance of that system. The real contribution of the leadership in a change period relies on managing the dynamics, not the pieces, as Duck (1993) explains. Main conclusions from a review of case studies dealing with change efforts The case studies reviewed have been selected to cover a range of different organizations that focus the change effort in different ways or by using different approaches: . efforts for lean production and development of value streams; . efforts for reengineering; and . efforts for change based on total quality approach. in order to assure that the conclusions of this review have a generic validity. Process Planning

Information and Communication Processes

Business Processes

Strategy

Structure

Processes

- Technology - New Organization Forms - Globalization

Figure 1. Strategic management framework

Culture Source: Hax et al. (1996)

- Processes - Activities

Control and Reward Processes

Performance - Benchmarking - Actitivity-Based Costing - Balanced Scorecard

The main conclusions of the review of the case studies are the following: . Visible and sustained top management commitment is a prerequisite for any change effort. . Human resources constitute a key parameter for any change effort and also a great obstacle, if not managed in a way that adequately addresses those issues of the change effort that affect people (potential for job loss, career path, requirements for new skills). . Internal communication should be planned to maintain focus and facilitate human resources management. . The new process-based teams or, in general, the new structures should be designed and implemented in a way that minimizes internal conflicts with the existing functions and does not lead to elimination of valuable function expertise. . The external environment should be carefully studied to avoid conflicts or discontinuity of the change effort. Review of literature addressing the supply chain performance The assessment of the performance of the supply chain or its sub-systems has been widely addressed in literature. In the late 1990s, the establishment of supply chain performance evaluation or quality measurement methods progressively moved from reactive to proactive approaches which had a process orientation and a high level of integration and focused issues related to the overall process management and corporate decision making. While the approach of Stewart (1997) structures the overall improvement planning of supply chain on a multilevel process analysis, beginning from four basic supply chain processes (plan – source – make – deliver), proposals of Fung et al. (1998) and Leonard et al. (2004) are based on fundamental concepts of total quality management, as a means to influence, first the strategic process and, second, business operations and also on a strong focus on customer satisfaction. Good coordination with key suppliers is reported by Aslanertik (2005) to increase product availability, which is crucially important for customer satisfaction. Beamon et al. (1998) emphasize the importance of process quality measurement across the supply chain system in the “plan-do-check-act” framework of total quality management and, from another point of view, Sherman (1998) considers the supply chain as one of the demand management structures, and studies the optimization of its dynamic interaction with demand and value chains. Logistics performance measurements based on hierarch models have also been suggested by Rafele (2004).

Methodology To develop the overall approach for the SCTP and the relevant management tool, the following concepts are adopted as the focal ones: . The strategic management framework (Shtub et al. (1994)), which determines the strategic planning process of an organizations. In order to be successful, any kind of change efforts should determine the strategy of the organization for a certain period of time.

Transforming the supply chain

851

JMTM 17,6

.

.

852

.

.

The significance of the proper implementation of human resources management techniques, since the human resources parameter is approved to be maybe the more sensitive one, especially during change periods. The organizational structure and culture that has to be in sequence with the priorities and concepts of the change effort. The marketing integration throughout the overall organization, as principle for a customer oriented supply chain. The approach of quality evaluation of all steps of the transformation process.

The proposed tool is the result of the study. It includes a modular part, to deal with the distinct supply chain processes, giving organizations the opportunity to use the tool in a way and extent suitable for their special characteristics and needs. The transformation process can be successfully managed as a project, taking under consideration the major framework of project management as it has been developed by Shtub et al. (1994) and the analysis and conclusions mentioned in the section Background of the paper. So the final structure of the output of this paper has a form of a flexible methodological project management tool. Presentation of the SCTP tool – analysis Figure 2 shows the proposed approach of the SCTP, in the form of a flexible methodological project management tool. Detailed guidance for each project phase and a set of criteria of successful performance for carrying out an on-going quality evaluation procedure are presented accordingly. Phase 1 – identifications In this phase, the project of SCTP becomes as specific as possible, considering its scope and the approach that has to be adopted. The top management examine the overall business environment (internal/external factors) and plan the new supply chain strategy within the strategic management framework. SCTP begins from the determination of the transformation objectives set by the organization. Such objectives are mostly related to the critical parameters for the development of competitive advantage: cost reduction, total quality improvement, differentiation, and they reflect the special characteristics and needs of the organization. Distinct supply chain processes: order processing, procurement, inventory management (raw materials – in process – finished goods), warehousing, transportation, distribution, customer service and reverse logistics are rated against the transformation objectives, taking under consideration the following criteria: . How direct is the impact of the process on the final customer. . The automation level of the process. . The impact of the process on the operational cost. . The existing rates of the quality indices of the process. The significance rate of each criterion and the rating scale are determined by the organization, according to, among other parameters, the overall corporate performance objectives. Figure 3 shows an indicative sheet to be used for rating the supply chain processes against a number of possible transformation objectives.

Transforming the supply chain

PHASE 1 – Identifications Project content: Transformation Objectives

Human resources needed – Project team

Overall corporate performance objectives

Rating of supply chain processes against transformation objectives

Selection of the pilot project

853

Top management commitment New Supply Chain Strategy

PHASE 2 – Analysis of Supply Chain Processes / Framework for process modules

PHASE 3 – Re-examine the Selected Pilot Project / Preparation of an Integrated SCTP Project Plan / Implementation and Evaluation of the Pilot Project

PHASE 4 – Realization of the Transformation Process Step 1

Modular process plan

Step 2

Project quality evaluation and overall performance objectives review

Step n New Customer Focused Organizational Culture

PHASE 5 – Project Completion / Results Evaluation

The output of the rating procedure gives the essential guidance to be used for the following phases of the transformation process. More specifically, the existence of similar ratings in the rows of the sheet identifies the priorities of the transformation process, whereas the column of the sheet with the higher total rating score identifies the process that should be selected for planning the pilot project. The issue of the human resources requirements for such a project is linked with the output of the rating procedure. Actually, these requirements cannot be described in general, but a SCTP project team and a project leader are necessary, to run the overall

Figure 2. An initial approach of a flexible methodological TP project management tool

Figure 3. Indicative supply chain processes rating sheet

Creating Customer Experience People Involvement Increase

Simplicity Increase

Transformation Objectives ↓ Process Integration

Order Processing

Procurement

Inventory Management

Warehousing

Transportation

Distribution

Customer Service

854

SC Processes →

Reverse Logistics

JMTM 17,6

project and assure people involvement and the needed communication and training. It is important to mention that the SCTP project team must have the skills to act as leaders for the overall effort, not simply as project managers. Additionally, during planning the transformation, the behavioural issues between the members of the top management should be managed with the help of a facilitator (external specialist or member of the SCTP project team). In any case, top management commitment has to be sustained and demonstrated. The SCTP project is proposed to be in full alignment with the overall corporate performance objectives, which reflect the corporate strategy. The reason is that a success factor of a transformation process is not to influence negatively the balanced operation of the organization and to maintain flexibility, during its progress. Another reason is that the people of the organization should be kept focused on performance results. A third reason is that the overall supply chain process is cross-functional and covers the total flow of both materials and information. So, in the phase of identifications, the overall corporate objectives have to be set and so have the respective delegations for achievement, in interaction with the SCTP project team. Finally, a pilot project has to be identified, which means that the new approach has to be implemented to a selected process, expecting to lead to pre-specified results. As mentioned before, the selection and planning of the pilot project is based on the rating procedure. The evaluation of the pilot project is the criterion for the pre-assessment of the performance of the SCTP project, which is carried out at phase 3.

Transforming the supply chain

855

Phase 2 – analysis of supply chain processes/framework for process modules As phase 1 is completed, the SCTP project team should carry out a detailed analysis of distinct supply chain processes, to get the information needed for the transformation process. Taking under consideration that key business processes have, to some extent, specific inputs and expected outputs, improvement opportunities exist in the process value chain (Figure 4) and, consequently, in the way business process sequence is supported by organization structure, so the relevant analysis for supply chain processes has a determinative role in the SCTP project. The analysis of supply chain processes may result to new roles, new work positions, new authorities, reporting relationships and organizational interactions. In addition to that, this analysis determines the necessary framework for planning the modules for each of the distinct supply chain processes. Such a framework addresses the following areas (the list is not exhaustive): . new policies; . new activities and procedures; . different flow of information;

Business Process Sequence Inputs

Outputs

•

• Product/Service C • Value C

• • •

Materials A Non-materials B Value A Value B

A + B + Value added by business processes = C

Figure 4. Business process value chain

JMTM 17,6

.

. . .

856

different interface with the external environment (customers, suppliers, competitors, society); need for changing the sequence of processes; opportunities for integration, including marketing integration; and new performance measurement systems.

The output of phase 2 gives the necessary information and determines requirements for the preparation of the modular process plan and an integrated SCTP project plan, carried out at phase 3. Phase 3 – re-examine the selected pilot project/preparation of an integrated SCTP project plan/implementation of the pilot project In this phase, a re-examination of the suitability of the selected pilot project can be useful, taking under consideration the results of the supply chain process analysis. Before the preparation of the integrated SCTP project plan, the implementation of the pilot project has to take place and the results should be evaluated against pre-specified requirements. The output of the evaluation is a determinative factor for both the integrated SCTP project plan and the realization of the transformation process (phase 4). An integrated SCTP project plan includes: . Modular process plan: a set of planned changes in each of the distinct supply chain processes, leading to the transformed overall supply chain. To prepare the modular process plan, the framework mentioned above (phase 2) determines the areas of concern. . Prioritisation of steps of transformation. . Time schedule. . Budget. . Definition of SCTP project milestones where project quality evaluation should take place and improvement measures should be taken if needed. The criteria, against which the evaluation takes place, consist of the review of the transformation objectives and the overall corporate performance objectives (set at phase 1), and also those mentioned in the specific paragraph below. . The necessary training and communication activities, in order to assure that people in the organization will adequately support the new approach and participate in teamwork. The Integrated SCTP project plan may be modified to some extent, taking under consideration the results of the implementation of the modular process plan and the results of the SCTP project evaluation. Phase 4 – realization of the supply chain transformation process In this phase the transformation process evolves, facilitated by the integrated SCTP project plan, and the modular process plan is implemented throughout the steps of the transformation process realization. Although the SCTP project steps are determined according to the transformation process of the specific organization, in generic terms, they may include the following:

.

. . . . . . .

.

communication the new approach among people, in order to minimize their resistance to change; design of new organization structure (to the necessary extent); re-definition or re-engineering of work methods; development of cross functional teams; people training on the new roles and responsibilities; re-examination of business plans; development of differentiated relationships with suppliers; definition of specifications for new information technology tools and automations; and re-orientation of customer relationships and marketing plans.

At the defined at phase 3 SCTP project milestones: .

.

.

.

A review of the transformation objectives and overall corporate performance objectives fulfilment rate has to take place, as a horizontal progress control network. An evaluation procedure has to be implemented against the additional criteria proposed below. Feedback is taken from people in the organization, customers and suppliers, in relation to the activities indicated by the modular process plan, to assure their suitability. Feedback is taken from people in the organization, to assure that the new customer focused organizational culture is being developed and gradually adopted by them.

Phase 5 – completion of the SCTP project and final evaluation of results During this final phase, the overall results of the SCTP are evaluated by using the criteria proposed below. The SCTP project team reports the results to top management and communicates them to people. It should be mentioned that such transformation process may never end. What we mean here is that a pre-set level of change has been accomplished and the dynamics for a continuous transformation culture have been developed within the organization and among people. For this reason, final evaluation results give feedback for re-examining the modular process plan and also the initial phase of identifications of a re-transformation process, so a continual change loop is activated. Criteria of successful performance of the SCTP project Pre-assessment criterion Results of the pilot project. As it is mentioned above, by completing the pilot project, which means that a changed mode has been implemented to a selected process, the results are evaluated against pre-specified requirements. As an example, we can describe the case of an order-taking process redesign, expecting to lead to a saving of one work position and time minimization. By implementing the redesigned process, it can be seen if expected improvements are achieved and, as a result, the redesign method is (partially or fully) validated or not.

Transforming the supply chain

857

JMTM 17,6

858

In-progress evaluation criteria Consistency to project time schedule and budget. To get new situations (and the relevant reactions of people) matured and to continuously balance between the previous the next step of the process, the necessary time must be given and, respectively, budgeted. The consistency to the time schedule and budget is a fundamental criterion of successful performance for every project and also for SCTP project. Review of the transformation objectives and overall corporate performance objectives. The transformation objectives and the overall corporate performance objectives, which have been set at the initial phase of the SCTP project, should be reviewed, to assure consistency with corporate strategy and to enhance the results orientation of the effort. The results of this review provide feedback for a probable re-examination of the integrated SCTP project plan. Internal conflicts/duplication of activities. As the transformation process evolves, the project team should: . get involved in the decision making process of the operational level to detect internal conflicts; and . focus on examining procedures and work flows to check out cases of overlapping or duplication of activities and to plan corrective arrangements. Final evaluation criteria The Consistency to project time schedule and budget and Review of the transformation objectives and overall corporate performance objectives criteria are used also in the final evaluation of the SCTP project. Additionally, the following criteria are proposed: Measurement of productivity of selected work positions. The productivity of selected work positions which are directly linked to supply chain processes, is more sensitive to changes. This can be measured as an evaluation criterion of the change process. For example, the implementation of different inventory management methodologies may lead to significant changes for the production workforce. Measuring the productivity of the relevant work positions when the transformation process has taken place (and after an adequate amount of time has passed), gives feedback for evaluating the process. Impact on key suppliers/customers. Balancing the SCTP with the external parameters is an important factor of the whole effort. To evaluate this factor, both a key supplier and a key customer of the organization can be selected as focal points, to detect the impact of changed work methods on the overall relationship between them and the organization. Customer satisfaction rate. The improvement of customer satisfaction rate is a key objective of any SCTP of an organization. In the phase of the final evaluation of the SCTP project, the customer satisfaction rate should be measured and compared to that of the period before the change effort, to provide additional feedback for evaluating the transformation process, in relation to the competitiveness increase achieved. Pilot test: the experience of the implementation of the tool to a big organization The SCTP project management tool proposed in this paper has been implemented to a big organization in Greece, which designs, produces and distributes furniture. The organization was founded in 1980 and had an impressive development during the last two decades. It has become the leading firm in the Greek market of furniture since

1995. The number of employees is about 700 (including the workforce of the factory). The products are distributed through a network of franchising shops in Greece and they are also exported to countries of the European Union and North America. The organization implements an ISO 9001 certified quality management system and an ISO 14001 certified environmental management system. Considering the changing market context and the increasing competition, the need for a more flexible and cost-effective supply chain structure had been initially identified. Following the phase 1 (Identifications) of the tool the following results were reported: . The production planning function should be transformed to a supply chain planning function, to manage the inventories of materials and products in an integrated way. . Since, the activities of the order receiving department did not actually add value, due to the fact that orders could be automatically transferred from shops to the new supply chain process planning function, the role of this department should be either transformed or eliminated. . Quality and environmental management systems should be integrated and unified procedures should be planned in relation to the key supply chain processes. . Customer service activities should be redesigned in order to achieve quicker responsiveness. The implementation of the tool led to the development of an integrated supply chain function, which would integrate planning, purchasing, production, warehousing distribution and transportation processes. The customer satisfaction rate had an increase of 25 percent, six months after the fulfilment of the transformation and three key suppliers reported a significant decrease of the duration of the overall purchasing process. Conclusions By reviewing change efforts that have taken place in big organizations, it can be seen that the success of such efforts depends on a number of correlated parameters, among which, those of top management commitment and behavioural issues of human resources are the most critical. Additionally, conflicts between the previous organizational structure and the one to be developed and also between the internal organizational transformations and external environment alarm the need for the overall effort to be systematically managed by a project management tool. The tool proposed in this paper for the transformation of the supply chain, provides a flexible sequence of project phases, linked together in a loop and continuously assessed against specific evaluation criteria, to assure the project quality and the continual transformation dynamics. The tool is structured in a way that allows a high level of customization to take place, considering its implementation to organizations of different sizes, industries or cultural approaches. So there are not any particular limitations in the exploitation of the tool to different organizations. An advantage of the tool is that incorporates the overall performance objectives of the organization (which reflect its strategic framework) in the initial phase of SCTP

Transforming the supply chain

859

JMTM 17,6

860

project identifications and uses the review of these objectives as a horizontal progress control network for the project evaluation. Another important issue is that a modular process plan is included in the tool to facilitate the transformation effort during all steps of its realization in a flexible way. Areas for further research may include the identification of guidelines for more extended arrangements in relation to the communication plan (since such a plan is the basic human resources management platform which is necessary for any change effort) and the development of an intranet platform to support the implementation of the tool. References Aslanertik, B.E. (2005), “Model-supported supply chains for cost-efficient intelligent enterprises”, Journal of Manufacturing Technology Management, Vol. 16 No. 1, p. 86. Beamon, B.M. and Ware, T.M. (1998), “A process quality model for the analysis, improvement and control of supply chain systems”, Logistics Information Management, Vol. 11 No. 2, pp. 105-13. Chandler, A. (1962), Strategy and Structure, MIT Press, Cambridge, MA. Drucker, P. (1988), “The coming of the new organization”, Harvard Business Review, January/February, pp. 45-53. Duck, J.D. (1993), “Managing change: the art of balancing”, Harvard Business Review, November/December. Fung, P. and Wong, A. (1998), “Case study: managing for total quality of logistics services in the supply chain”, Logistics Information Management, Vol. 11 No. 5, pp. 324-9. Hax, A.C. and Majluf, N.S. (1996), The Strategy Concept and Process: A Pragmatic Approach, Prentice-Hall, Englewood Cliffs, NJ. Leonard, D. and McAdam, R. (2004), “Total quality management in strategy and operations: dynamic grounded models”, Journal of Manufacturing Technology Management, Vol. 15 No. 3, p. 266. Martin, R. (1993), “Changing the mind of the corporation”, Harvard Business Review, November/December. Rafele, C. (2004), “Logistics service measurement: a reference framework”, Journal of Manufacturing Technology Management, Vol. 15 No. 3, p. 290. Sherman, R.J. (1998), Supply Chain Management for the Millennium, Warehousing Education and Research, Oak Brook, IL. Shtub, A., Bard, J.F. and Globerson, S. (1994), Project Management: Engineering, Technology and Implementation, Prentice-Hall, Englewood Cliffs, NJ. Stewart, G. (1997), “Supply-chain operations reference model (SCOR): the first cross-industry framework for integrated supply-chain management”, Logistics Information Management, Vol. 10 No. 2, pp. 62-7. Corresponding author Evangelia D. Fassoula can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints