International Journal of Computer Integrated Manufacturing Vol 23, No 1, January 2010, 1–19 Development of collaborative transportation management framework with Web Services for TFT–LCD supply chains M.-C Chena*, C.-T Yehb and K.-Y Chenc a Institute of Traffic and Transportation, National Chiai Tung University, Taiwan, ROC; Department of Industrial Management, National Taiwan University of Science and Technology, Taiwan, ROC; c Department of Industrial Engineering and Management, National Taipei University of Technology, Taiwan, ROC b (Received 20 July 2008; final version received May 2009) Under the fierce global competition, enterprises face the challenge to respond quickly and accurately to customers’ diverse requirements Excessive lead time, improper control of transportation resources and inaccessibility of tracking information may lead to ineffective and unreliable delivery The seamless integration of trading partners such as suppliers, manufacturers and global third party logistics service providers (G3PLs) can improve the supply chain execution With supply chain collaboration, lead time can be significantly shortened whereas the resource utilisation can be highly improved Recently, collaborative transportation management (CTM) has provided the collaborative mechanisms of information sharing and order fulfillment for carriers and trading partners in supply chains CTM initiative can reduce the ineffective transportation sections and better the delivery effectiveness The heterogeneous information systems first should be integrated for starting up information sharing and data integration among carriers and trading partners Web services possess characteristics of flexibility and interoperability that are suited for developing the inter-enterprise collaboration platform by integrating heterogeneous systems With the web-services based CTM (WS–CTM), carrier and trading partners can collaborate on the process of order fulfilment In this paper, the WS–CTM framework is developed to collaboratively manage transportation and distribution for a supply chain of thin film transistor liquid crystal display (TFT–LCD) The proposed WS–CTM can assist the panel manufacturers, system manufacturers and G3PLs to reduce the uncertainty in distribution, and improve the supply chain performance Keywords: supply chain management; collaborative transportation management; global logistics; Web Services; TFT–LCD; e-Business Introduction With the fierce global competition and the falling profit margin, most companies engage in a global supply chain to maintain the market share and to intensify the profit (Tyan et al 2003) Managing the global supply chain is much more complicated compared with managing the domestic supply chain (MacCarthy and Atthirawong 2003) Excessive lead time, improper control of transportation resources and inaccessibility of tracking information may lead to ineffective and unreliable delivery The seamless integration of trading partners such as suppliers, manufacturers and global third party logistics service providers (G3PLs) can improve the supply chain performance Generally speaking, lengthy lead time results in a higher level of inventory and a higher cost In addition, the global distribution causes a notable increase of transportation cost (Meixell and Gargeya 2005) The decision making in a global supply chain has a higher impact on performance Collaboration and strategic alliance are *Corresponding author Email: ittchen@mail.nctu.edu.tw ISSN 0951-192X print/ISSN 1362-3052 online Ó 2010 Taylor & Francis DOI: 10.1080/09511920903030353 http://www.informaworld.com known as effective mechanisms reducing the uncertainty and risk in global business For achieving a certain customer service level, suppliers generally concern about the accurate and reliable delivery of products to the locations assigned by buyers On the other hand, 3PLs focus on satisfying the delivery requirements set by customers and maximising the utilisation of transportation resources To maximise the supply chain performance globally, involved members need to collaborate on planning, forecasting and execution Collaborative planning, forecasting and replenishment (CPFR) has hence been introduced in supply chain management Traditionally, the related issues of supply chain collaboration centre around material supplier–manufacturer and manufacturer–retailer The logistics of order fulfilment is an extremely complicated process which includes order generation, order picking, shipping, transportation, payment, invoicing and so on In particular, these activities are performed geographically distributed in global business Enterprises, M.-C Chen et al generally, outsource their logistics function to 3PLs to reduce the logistics cost as well as to concentrate on their core competitive advantages Carriers or 3PLs are not often considered in the seller–buyer collaboration and strategic alliance However, carriers play the role of order execution as physically delivering the goods to the locations appointed by receivers In addition to CPFR, collaborative transportation management (CTM) (CTM Sub-Committee of the VICS Logistics Committee 2004) provides a collaboration mechanism among seller, buyer and carrier The initiative CPFR focuses primarily on the planning of order fulfilment, whereas CTM makes up the link between planning and execution For bringing the carrier in collaboration, shippers (sellers) and receiver (buyers) need to share the related information such as order forecasting, replenishment plan, etc such that carriers can plan the transportation resources in advance and execute the delivery better Sellers and buyers can easily track the goods in transit with the mechanism of distribution information sharing The exceptions in goods delivery (e.g., delivery delay) can be controlled in real time as well as can be resolved collaboratively Information sharing is an essential task of supply chain collaboration However, the heterogeneous information systems among supply chain partners cause a huge difficulty in information exchanging and sharing This paper proposes a CTM platform (namely WS– CTM) for a supply chain of thin film transistor liquid crystal display (TFT–LCD) This platform integrates the capability of Web Services with heterogeneous systems in supply chains as a virtual system Characteristics of Web Services such as flexibility and interoperability are particularly suited for developing the inter-enterprise collaboration platform among heterogeneous systems The proposed WS–CTM can assist the panel manufacturers, system manufacturers and G3PLs to reduce the uncertainty in distribution, and improve the supply chain execution They can collaboratively execute the order fulfilment on WS–CTM platform to increase the accuracy and reliability of distribution Web services and supply chains Web Service (WS) is a new information technology of web application (W3C 2004) Web Services are selfcontained, self-describing, modular applications that can be published, located, and invoked across the web They are bases on open standards, i.e hypertext transfer protocol (HTTP), extensible markup language (XML), simple object access protocol (SOAP), universal description discovery and integration (UDDI), web services description language (WSDL) and a common architecture, service-oriented architecture (SOA), to integrate heterogeneous business systems and to support interoperable machine-to-machine interaction over a network (W3C 2004) With the self-recitation property of XML and WSDL, various software components can recognise one another SOAP is a messaging protocol which allows components to interact each other UDDI is a set of protocols which can describe, register, search and integrate service components With the trend of global supply chain and door-todoor service, the traditional distribution architecture may not meet the diversified customers’ requirements In global business, inter-modal and/or multi-modal transportation are necessary to execute the order fulfilment process which increase the lead time in consolidation and transportation (Tyan et al 2003) The strategic alliance among sellers (shippers), buyers (receivers) and carriers (3PLs) can foster the logistics efficiency and quality (Andersson 1995) The trading partners can benefit from the strategic alliance with 3PLs for performing distribution activities The benefits include the larger economic scale, larger bargaining power, better service accessibility, enabling knowledge learning, investment reduction and so on G3PLs serve as a virtual global distribution centre to link the supply chain members globally dispersed (Tyan et al 2003) Taking Taiwan’s Note Book (NB) industry as an example, although companies have a superior capability in manufacturing, they integrate G3PLs (e.g FedEx) to distribute products for enabling door-to-door service Web Services support a new model for inter-system and inter-enterprise collaboration Web Services can realise the network manufacturing, in which heterogeneity exists and must be addressed Shen et al (2007) proposed ontology to deal with the heterogeneity in Web Services composition Hung et al (2005) proposed a Web Services based e-Diagnostics Framework (WSDF), which integrates diagnostics information with Web Services technologies WSDF can automatically collect equipment data, remotely diagnose, fix, and monitor equipment, and analyse and predict the equipment performance over the intranet and internet Dhyanesh et al (2003) proposed a methodology for constructing Web Services based infrastructure for cross-enterprise collaboration, namely DEVISE, which consists of a set applications Flexibility, efficiency and trustworthiness have become the crucial concerns in the fierce market, and Web Services can provide a solution for these concerns and business processes integration (BPI) Yang et al (2005) developed a trustworthy Web Services based framework for BPI With the rapid growth of e-commerce, business competition is much fiercer than years ago A company needs to build stronger relationship with its trading International Journal of Computer Integrated Manufacturing partners and customers for keeping their competitive advantages E-business is to perform the core business processes on the internet which not only include purchasing, selling products and services, but also collaborating with trading partners on the internet The process integration is the key for successful e-business implementation Chen et al (2007) proposed a collaboration architecture that uses Web Services for Business Process Management (BPM) in support of collaborative commerce (C-commerce) With the advancement of Web Services and BPI tools, BPM can execute Ccommerce more flexibly and dynamically For modern enterprises, various kinds of systems and applications need to interoperate more flexibly and to be easily integrated Kalogeras et al (2006) presented a distributed architecture utilising Web Services as a single common interface to vertically integrate the application systems Michalakos et al (2005) used Web Services to support the CPFR implementation, especially to facilitate the process of exchanging forecasting outputs among heterogeneous systems of trading partners Web Services provide a mechanism for the integration of buyer’s and seller’s heterogeneous systems to facilitate information flow and collaboration along a supply chain Supply chain collaboration has become an important issue to enterprises Supply chain partners can benefit more from the higher degree of collaboration (Simatupang and Sridharan 2004, Holweg et al 2005) Supply chain collaboration aims at integrating all partners to work as one virtual network toward common goals (Mentzer et al 2000) Because retailers have a greater power in supply chains, CPFR Table The applications of TFT–LCD Panel size (in.) Applications 5–7 Mobile Phones, PDAs, Handheld PCs, Navigators, other portable information and communication products Sub–NB PCs, small LCD TVs NB, PCs LCD monitors LCD monitors, LCD TVs 7–10 10–14.1 15 Over 15 Figure The TFT–LCD supply chain architecture programs are frequently initiated by retailers They therefore play the role of hub in supply chains in order to reduce the bullwhip effect From Chen et al (2007), it may be better to start the collaboration initiative from a retailer (buyer)-driven program CTM is an initiative of deeper and wider supply chain collaboration because it extends the procedure of CPFR, as well as CTM invites more partners to join the initiative CTM transforms order forecasts generated by CPFR into shipment forecasts, and it brings about collaboration among shipper, receiver and carrier to ensure accurate order fulfilment (Esper and Williams 2003) The TFT–LCD supply chain Thin film transistor liquid crystal display (TFT–LCD) industry is a technology-, capital-, and skilled personnel-intensive one in which the products have the characteristics of short life cycle, high cost and high value-added The applications of TFT–LCD are summarised in Table (ITRI 2002) Owing to the governmental initiatives, Taiwan has become one of the main producers of LCD monitor The competitive advantage of Taiwan’s TFT–LCD industry is the huge demand stemmed from the manufacturers of NB and Personal Computer (PC) Owing to the complex product structure and the highly geographically dispersed component providers, the supply chain integration can be a critical basis for the success of TFT–LCD industry Excellent logistics network and collaboration can support a seamless and effective integration for a TFT–LCD supply chain The TFT–LCD supply chain is complex with a wide variety of engaged members belonging to several tiers Figure schematically illustrates the supply chain architecture of TFT–LCD consisting of materials suppliers, panel manufacturers, module and system manufacturers (e.g., LCD monitor, LCD TV, PDA, cellular phone), brand-owned channels Taiwan’s TFT–LCD industry focuses on the section of manufacturing (ITRI 2006) The current difficulties in this section are as follows: (a) Materials suppliers: The demand of materials is difficult to control because the downstream M.-C Chen et al module/system manufacturers’ demand variations are high Additionally, the complexity, long lead-time and information invisibility in the TFT–LCD supply chain cause the bullwhip effect such that the inventory level is set high to buffer the sharp fluctuation of production plan Figure The international supply chain of Taiwan’s TFT–LCD industry Figure The global logistics network of Taiwan’s TFT–LCD industry International Journal of Computer Integrated Manufacturing (b) Panel manufacturers: For panel manufacturers, the order cycle time of material procurement is extended because the critical components are Figure The 14 steps of CTM Figure The WS–CTM framework generally geographically dispersed Generally, system manufacturers not share their demand information to panel manufacturers The panel demand is difficult to forecast due to the information invisibility and the high demand variation In such a situation, the safety level of inventory is high resulting in high inventory risk and cost The production plans of panel manufacturing may change frequently over time since most of the panel manufacturers mix-up the production with OEM (Original Equipment Manufacturer), ODM (Original Design Manufacturer) and OBM (Own Branding and Manufacturing) Furthermore, the panel is commonly manufactured in multi-factory which causes the difficulty in production planning and the increase in transportation distance and cost (c) System manufacturers: The business model of system manufacturers as well mix-up with OEM, ODM and OBM Since system manufacturers and panel manufacturers not have close partnership, system manufacturers can not flexibly respond the customer demands with high fluctuation To encounter such situations, system manufacturers raise the inventory level for buffering, and raise the purchasing price for higher supply priority The inventory cost and purchasing cost therefore increase considerably M.-C Chen et al The materials of panel manufacturers in Taiwan are supplied by local suppliers and foreign suppliers Additionally, the increasing offshore manufacturing of LCM and system manufacturer rises up the difficulty of the transportation of large-size panels (ITRI 2006) After the final products have been assembled by system manufacturers, they are delivered to channel companies and brand companies which are also geographically dispersedly as shown in Figure In the TFT–LCD supply chain, from suppliers through panel manufacturers and system manufacturers to channel companies and brand companies, the cross-country transportation is required which results in higher logistics complexity and longer lead time (refer to Figure 3) The panels can be used in various applications For example, 17’’ panels may be delivered to LCD monitor manufacturers, whereas 5’’ panels may be delivered to mobile phone manufacturers Therefore, the logistics from panel manufacturers to system manufacturers is outsourced to 3PLs for specialty concern However, the shipment forecasts are not shared to 3PLs by TFT–LCD manufacturers causing frequent LTL (Less Than Truckload) transportation and higher transportation cost As mentioned above, it is necessary to effectively integrate the TFT–LCD manufacturers with 3PLs, particularly G3PLs, to reduce the transportation cost and to satisfy the service level The information such as order forecast and shipment forecast of panel manufacturers and system manufacturers can be shared to G3PLs for better transportation planning and more reliable transportation CTM proposed by Voluntary Figure The global vision of WS–CTM framework Inter-industry Commerce Standards (VICS) can be developed as a platform between manufacturers and carriers for collaborative transportation In addition, Web Services can smoothly enable the development of cross-enterprise collaboration initiative Development of WS–CTM 4.1 Collaborative transportation management Collaborative transportation management (CTM) is defined by VICS (CTM Sub-Committee of the VICS Logistics Committee 2004) as ‘a holistic process that brings together supply chain trading partners and service providers to drive inefficiencies out of the transport planning and execution process.’ CTM mainly aims at improving the interaction and collaboration between three major parties, a shipper (seller), a carrier (3PL), and a receiver (buyer) Either the shipper or receiver may be the owner of carrier under CTM The owner is responsible for hiring and paying for the transportation service CTM can be an extension of CPFR such that CTM extends the collaboration scope to shippers, receivers and carriers (3PLs) As both inbound and outbound transportation flows are included in the CTM processes, both the shipper and the receiver can perform some of the CTM steps, while other steps are performed individually by either the shipper or receiver The leading party is responsible for the carrier relationship/ contract and the CTM steps CTM process consists of order/shipment forecasting, capacity planning and International Journal of Computer Integrated Manufacturing scheduling, order generation, load tender, delivery execution, and carrier payment Participating parties collaborate on transportation by sharing the essential information of demand and supply about forecasts, event plans, expected capacity, etc., schemes and capabilities for enhancing the performance of the overall transport planning and execution process, and assets, where feasible (i.e., trucks, warehouses) CTM can be divided into three phases composed of 14 steps (refer to Figure 4) as follows (Browning and White 2000): (a) Strategic phase: This phase defines the strategic issues of collaboration and includes of the two steps, Develop Front-end Agreement and Create Joint Business Plan The first step consists of the owner of carrier, which products, locations and types of shipments are Figure The process and functions of WS–CTM included in the collaboration, the exception management plan, and a summary of key performance The second step involves the aggregate planning phase, in which planned shipment volume should be matched to equipment asset plans (b) Tactical phase: It defines the procedure of shipment planning beginning with the generation of a product/order forecast, and ending with the generation of shipment forecast In the step of order/shipment forecast, the shipment forecast should be created according the collaborative scenario The next two steps are to identify the exceptions for order/shipment forecast and resolve exception items of order/shipment forecast based on the exception management plan (c) Operational phase: This phase defines the procedure for order execution and fulfilment M.-C Chen et al including of the shipment tenders, identification and resolution of exceptions for tenders forecast, freight contract confirmation, delivery, invoice and managing performance CTM extends CPFR to the order execution stage by firstly translating the order forecasts generated from CPFR to shipment forecasts Except sellers and buyers, carriers join CTM to play the role of order shipment From the study in Esper and Williams (2003), with the help of information technology, CTM improves the operations and efficiency of all members joining the collaboration In 2000, Wal-Mart extended the CPFR initiative with Procter & Gamble to a partial CTM by integrating the transportation logistics company, J.B Hunt (Dutton 2003) Although this project is a partial CTM, two of these three partners Figure The use case diagram of Identifying and Resolving Delivery Exception Figure obtain noteworthy benefits For Wal-Mart, the number of steps to process goods for promotions is reduced With sooner information exchanging, J.B Hunt can take action on information some days earlier than regular Therefore, J.B Hunt reduces the unloading time by 16%, and empty miles by 3% Procter & Gamble has no change in its outcomes from this CTM project 4.2 System analysis The key objective of CTM is to reduce the uncertainty in demand, supply and transportation through effective information sharing and collaboration For integrating sellers, buyers and G3PLs, a Web Services based CTM, namely WS–CTM, is developed in this paper The proposed WS–CTM can effectively respond the demand and improve the resource utilisation of transportation The WS–CTM platform is designed according to the process of CTM proposed by VICS (CTM Sub-Committee of the VICS Logistics Committee 2004) and users’ requirements The proposed WS–CTM as shown in Figure integrates the seller’s Enterprise Resource Planning (ERP) system, buyer’s ERP and carrier’s Logistics Management System (LMS) Through WS–CTM, partners can share the necessary information such as forecast, order, shipping, transportation capacity, etc, and they can then collaborate on order fulfilment The global vision of WS–CTM (refer to Figure 6) with multi-seller, multi-buyer and multi-carrier can be easily achieved through the WS architecture That is, the trading partners in global supply chain can communicate with each other through WS–CTM platform by using the internet The process and functions of The sequence diagram of Identifying and Resolving Delivery Exception International Journal of Computer Integrated Manufacturing Figure 10 The schema of WS–CTM relational database WS–CTM are illustrated in Figure Each function of WS–CTM is established based on CTM steps Note that the steps regarding the exception identification and resolving are integrated into a single process because the resolution of exception logically needs to be continued until no exception exists The developed platform mainly includes WS–CTM management platform, seller side CTM–ERP interface, buyer side CTM–ERP interface and carrier side LMS Taking the step of delivery exception identification and resolving as an example, the use case diagram illustrated in Figure displays that the client can 72 Z Cui and R Weston case 1, case and case Couples studied under (1), (2) and (3); then develop ‘to be’ versions of manufacturing systems (with alternative reconfigurations of P/R couple and related work flows) Having gained confidence in the models and key understandings about realisable behaviours from different model configurations, during Step D new reasoning can be developed about potential advantages and disadvantages of restructuring the models For example, restructuring might be made in conformance with a relevant manufacturing paradigm such as Lean, CM, RMS and DPD Whereas when using the Model-Driven Method to conceptually design a ‘green field’ EoSS manufacturing system it was proposed that an initial conceptual EoSS manufacturing system design would be developed by referencing common manufacturing architectures or by observing similar industrial system solutions, following which Steps A to D would be completed in an appropriate order The overall aim therefore has been to systemise experimental design and related conceptual thinking so that when simulating the operation of EoSS manufacturing systems it becomes possible to cope with the high levels of experimental complexity commonly found when computer executing models In this way early modelling reported in this paper was aimed at understanding behaviours of manufacturing systems used by collaborating ME’s working with the authors Following which specific and generalised observations were made about the impacts of different forms of Work Dynamic on the behaviours of alternative EoSS Figure Three main product categories of BGF company configurations of P/R couples By so doing specific recommendations for improvement were made to collaborators and generalised academic contributions could be made Case study 5.1 Company overview and enterprise model The case study company chosen was a UK based privately owned Small to Medium-sized Enterprise (SME) called BGF BGF makes-to-order household furniture by applying a push synchronisation policy BGF’s production sections include: a ‘machine shop’, an ‘assembly shop’; and a ‘spray shop’ Having collated customer orders for up to weeks, BGF creates and releases (twice a week) a production order (termed a ‘picking list’) to the production shops At this point in time the assembly shop coordinator requests parts from the machine shop The machine shop’s job list is continually forecast, and prioritised for the purpose of maintaining enough supply to the assembly shop It follows that ‘customer orders’ codeup a product dynamic (as earlier defined) while ‘production orders’ (or picking lists) constitute a Work Dynamic (which is an attenuated function of the coded product dynamic) BGF manufactures circa 350 different products (Rahimifard and Weston 2007); where the company divides these products into three main categories based on their functional differences, namely ‘cabinets’, ‘tables’ and ‘chairs’; see Figure To capture the process network of the case company the authors chose to deploy the Computer International Journal of Computer Integrated Manufacturing Integrated Manufacturing Open Systems Architecture (CIMOSA) Enterprise Modelling (EM) approach CIMOSA was specified and developed in the late 1980s and early 1990s (AMICE 1993) This choice was made because CIMOSA is an ISO Standard and has proven effective in many MEs of different sizes (Vernadat 1993) as a way of creating a process oriented big picture of any given ME which can be formally decomposed into interrelated parts that can be modelled at a level of abstraction that is determined by the modeller (Weston et al 2003, Chatha et al 2007) CIMOSA provides a consistent architectural framework for helping companies engineer change to their processes and resource systems This can lead to competitive advantage in terms of price, quality and delivery time (Vernadat 1993, 1996) The authors and their research colleagues in the MSI Research Institute visited the case company many times to interview its senior and middle managers This allowed them to formally capture process-oriented knowledge held collectively by those managers, and to organise this into a CIMOSA process-oriented requirements model The EM produced comprised a number of: context diagrams; interaction diagrams; structure diagrams; and activity diagrams These graphical models enabled MSI modellers to obtain sufficient knowledge about how the material and products pass through BGF processes at runtime This provided foundation knowledge to begin to create simulation models of the various production shops of BGF, with the purpose of supporting ME prediction and related business, manufacturing and logistical decision and policy making Figure Context diagram of BGF 73 Figure shows a context diagram of BGF, which was used as the starting point when creating CIMOSA models of the ‘make and deliver to (aggregated) order’ product realisation processes realised by BGF and its supply chain partners This figure shows that there are a number of supply chain companies and departmental units (which are defined as being domains (DMs) that collectively interoperate to achieve this process network For these independent units, the modeller can decide whether to model that domain in detail using CIMOSA modelling constructs, or to essentially treat that domain as a black box In the overall BGF study, prime concentration was on two domains, namely DM6 ‘Business management’ and DM7 ‘Produce and deliver furniture’ In this paper, focus of attention is almost solely on DM7 List of domains: DM1 Stockists: These are retailers of home furniture, who purchase products from BGF and other furniture manufacturers and supply them to furniture shops and major stores mainly throughout the UK DM2 Raw Material Suppliers: Suppliers of BGF raw material, mainly pine wood (from Scandinavian countries) but also paint, glue and leather DM3 Sub Products (e.g Chairs) Suppliers: Some sub-products are outsourced by BGF, mainly chairs that are part manufactured so that BGF can carry finishing operations and re-sell them at a higher price DM4 Miscellaneous Fixture Suppliers: Provide BGF with small items such as screws, hinges and door knobs DM5 Technology Vendors: Suppliers of machines, software and other utilities that require support maintenance 74 Z Cui and R Weston DM6 BGF Business Management: This internal domain of BGF obtains and processes orders, manages interactions with suppliers and customers and carries out business and financial transactions DM7 BGF Produce and Deliver Furniture; This internal domain of BGF realises the case companies’ ‘order fulfilment processes’ This domain is the main subject of study in this paper Having created the context diagram shown in Figure 9, the subsequent enterprise modelling steps centred on using CIMOSA constructs to capture a significant number of interaction diagrams; structure diagrams and activity diagrams The purpose of creating these diagrams is to decompose the targeted domain to enable the organised capture of knowledge from key knowledge holders operating at all levels in the domain By so doing the network of processes owned by that domain are explicitly described at all appropriate levels of modelling granularity (or detail) In the case of BGF, DM ‘Produce and Deliver Furniture’ and DM ‘Business Management’ were treated as CIMOSA conformant domains, and as such they were modelled in detail using CIMOSA modelling principles Figure 10 shows the high level structure diagram created to model DM This figure shows that there are sub processes within DP7, namely, business process (BP) 71 make furniture to order, BP72 spray and finish furniture production, BP73 package and deliver furniture, BP74 support and implement product introduction and BP5 manage and maintain production and transport capabilities Partly as direc- Figure 10 Structure diagram of DP7 ted by BGF managers, the research of the authors has been focused on the assembly manufacturing system of BGF, the operating requirements of which were described by BP71 There are sub BPs and enterprise activities (EA)s within BP71 Part of BP 71, namely sub BP 71-2 specifies how BGF assembles carcasses and fits components This is the specific focus area of concern of this paper The ultimate aim of CIMOSA modelling and decomposition is normally to create a network of dependent activity diagrams which are essentially process models that are formally linked together via higher level context, interaction and structure diagrams previously produced and described Figure 11 shows an example decomposition of production process and sequences within BP71-2 assemble carcass and fit components that were modelled at various levels of granularity in the form of CIMOSA activity diagrams The assembly shop of BFG comprises three sections, namely, table, cabinet and bed assembly sections Essentially, it was found that there are two main table types that BFG assembles They are: Drop Leaf (DL) tables and Farm House (FH) tables While product sub-groups within the cabinet area consist of Base units; Solid doors; Drawers; Rothley units; Display unit tops; Breakfront bases In the area of bed making is the Beau manor bed Operational processes and related activity sequences needed to realise some of these different product types are shown in Figure 11 In effect, Figure 11 illustrates the decomposition of BP71-2 at the enterprise activity International Journal of Computer Integrated Manufacturing Figure 11 Activity diagram of BP71-2 75 76 Z Cui and R Weston level such that it codes up the processing routes for different products as they are processed by the assembly manufacturing system of BGF The reader can observe that the ‘static process differences’ occur because of the PV in the system and that these processing differences can be clearly shown using the CIMOSA modelling technique However, the impact of other Product Dynamic factors cannot be explicitly described using standard CIMOSA modelling constructs, particularly where dynamic state changes need to be detailed It follows that in the case of BGF, CIMOSA requirements modelling was used to describe relatively enduring (structural) relationships between different elements of the BGF enterprise However to develop time-based behavioural models of the manufacturing systems used by BGF, so as to enable more effective enterprise engineering, it was necessary to transform the CIMOSA activity diagrams into either discrete event or continuous simulation models that can computer execute, and thereby model states and state transitions of modelled manufacturing systems, as different forms of Product Dynamics are input to these models 5.2 Simulation model development Simulation modelling can support ME design and change in a number of ways including: a) understanding required manufacturing system behaviours during the life phases of products; b) identifying possible solutions when ME change is required; c) visualising change in manufacturing system behaviours to facilitate prediction during planning These understandings led to the selection of two types of simulation technologies for use in this research, namely ‘discrete event simulation’ and Table Operation and op-time summary of selected table types of BGF (data source MSI) Collect from racking FH1 FH2 FH3 FH4 FH5 FH6 FH7 DL1 DL2 DL3 DL4 DL5 DL6 DL7 DL8 ‘continuous simulation’ In this paper though, discussion only describes use of a specific discrete event simulator, namely ‘Simul8’ (Simul8 2000) Here it was observed that discrete event simulation technologies are well suited to modelling some of the core variables of concern Simul8 was chosen mainly because of its availability and its provision of excellent technical support within the MSI Research Institute This is not claimed to be the best in class tool for this purpose but it has proven to be flexible and effective The second simulation technology chosen was a continuous simulation tool called ithink (ISEE 2007) which was designed to directly computer execute causal loops of manufacturing systems and their related business environments ithink deploys numerical integration and its use is well proven for large scale, complex systems modelling Complementary ithink modelling of the authors and their research colleagues is in support of supply chain and value stream analysis, but these topics are not covered in this paper A detailed study was made of the assembly operations carried out in the BGF table assembly section to assemble DL and FH table products Table illustrates some of the time study results obtained, which shows operation times taken by operators, as (aided by fixtures, tools and general purpose wood-making machines) they produce and assemble different table products Table also shows that there are seven different types of FH table and eight different types of DL table This shows that the sequence of assembly operations, and average operation times, are the same for all members of the FH table class Similarly, all DL table types share a common operation sequence, but some operation time differences can be observed within this product class As explained in section 5.1 and shown in Figures 10 and 11, prior to time study of DL and FH assembly 5 5 5 5 5 5 5 min min min min min min min Fit piano hinges Shape top Sand top * * * * * * * 15 15 15 15 15 15 15 15 * * * * * * * 20 20 20 20 20 20 20 20 * * * * * * * 10 10 10 10 10 10 10 * Assemble under-frame 20 20 20 20 20 20 20 10 10 10 10 20 20 20 min min min min min min min * Fit under-frame to top 15 15 15 15 15 15 15 15 15 15 20 20 20 20 min min min min min min min * Sand leg 15 15 15 15 15 15 15 20 20 20 20 20 15 15 min min min min min min min * Assemble under-frame 15 15 15 15 15 15 15 10 10 10 10 10 10 10 10 min min min min min min min International Journal of Computer Integrated Manufacturing sequences, BP 71-2 (Assemble Carcasses and Fit Components) and its constituent Enterprise Activities had been modelled using CIMOSA at a higher level of abstraction Here, BP 71-2 was modelled as part of the BFG ‘big picture’ process network Therefore the activity diagrams of Figure 11 and the table assembly sequences shown in Table fleshed out some of the detail of assembly process modelling, which had been hidden within the more abstract ‘in context’ model of BP 71-2 described by Figure 10 But the reader should also be aware that in this case the fleshing out was only partial as many other BFG product types (a total of around 350) have, as required by customers, to flow through the real BFG BP71-2 assembly process segment Furthermore, the reader should understand that BP71-2 is but one of many assembly processes realised in the assembly section of BGF When building simulation models therefore the model builder needs to cut through this complexity to enable creation and use of models of practical and effective simplicity while still being able to replicate actual behaviours of real ME segments The complexity of simulation models can grow very quickly also because not only the processing sub-system (and its complexity) needs to be modelled to exhibit system behaviours In addition, the human and resource system elements, their binding structures, and related work flows through process and resource elements need to be modelled However, the step-wise approach described in sections 5.1 and 5.2 of this paper illustrate how simulation model simplification and modelling building can in part be systemised In the real BGF Assembly Shop, DL and FH table types are but two of a total of 18 table product types produced Having considered particularly DL and FH table assembly as described by Figure 11 and Table 2, it was decided that BP71-2 could realistically be decomposed into six EAs This decomposition was Figure 12 77 coded in Simul8 by corresponding work centres, work entry and exit points, and queue modelling constructs; as shown in the screen shot of Figure 12 Each EA was studied in detail to develop groundwork knowledge needed to create this simplified simulation model which will be referred to as SM1 Essentially, SM1 was designed to simulate state behaviours of the Drop Leaf table assembly process Here, detailed study centred on: defining processing activities and routes; and on grouping activities into ‘roles’ that can be assigned to Work Centres (WCs) incorporated into SM1 Following which human and other resources needed to realise each WC were assigned Thereby SM1 was designed and developed to model the time-based behaviours of: enterprise activities EA 7.1.2.1 and EA 7.1.2.2 that are executed by WC1 Bench 1; EA 7.1.2.3 which is executed by WC2, which in turn involves CNC routing operations; EA 7.1.2.4 which is executed by WC3 to’ Sand tops’; EA 7.1.2.5, which is executed by WC4 and Fits under frames; EA 7.1.2.6 which is executed by WC5 and Sands legs; and finally EA7.1.2.7 which is executed by WC6, so as to assemble under frames It was observed that if the assembly process used to realise Drop Leaf tables is considered in isolation this constitutes a case system Therefore SM1 models a case EoSS manufacturing system This observation can be made as there are eighty types of DL table within the DL family, each of which requires a common manufacturing process so that essentially they share the same P/R couple In this example case EoSS system however, only relatively minor processing differences and resource requirements are needed for different DL types Having studied the product dynamic factors within the isolated DL table assembly system, it was observed that VD, PV and PMD (associated with any DL mix) can all impact on the table assembly system with potentially significant Simulation model of drop leaf table assembly process SM1 78 Z Cui and R Weston change in instantaneous values; which in turn will significantly influence the behaviours and performance of the assembly shop However, it was later observed that although the eight types of DL table are physically different in terms of dimension and feature, during assembly this only gives rise to a PMD mix that requires op time (i.e processing time) differences, which in turn requires relatively minor differences in resource requirements (compared to that found later when DL and FH table types were assembled together) As the factory needs to be able to assemble all DL types in uncertain quantities and mixes, it is however necessary to reprogramme the resource systems of the assembly shop (comprising both human and technical resources) every time different mixes of DL are assembled However, later it was observed that some of the DL tables have the same processing time and resource requirement, while others not Hence it was decided that the very similar DL table types could be grouped (into essentially a separate sub-class of table) Within this new group there were still dimensional (shape) differences between group members but this difference had previously been realised via make operations performed in the machine shop Therefore, with respect to operations performed in the assembly shop within this sub group, essentially there was zero variance (i.e PV ¼ and PMD ¼ 0) across members of that group; which in turn meant that the manufacturing system needed to assembly this sub-grouping of DL table products could reasonably be considered to be a case (economy of scale) manufacturing system (because only VD impacts as a significant product dynamic factor) Having made this observation, a new design of the DL table assembly system was proposed and simulated as SM1’ The conceptual design of SM1’ is shown in Figure 13 In the SM1’ system model, the input Work Dynamic was in effect reduced by conducting a GT Figure 13 analysis for product groupings; even though the original product dynamics caused by production order variation remained unaltered This meant that for the original product dynamics there were eight different types of DL table giving rise to variance, as individual entries to the system; but this was reduced to five by grouping products with the same processing time As a consequence differentiation between types of DLs was postponed to a downstream processing activity (namely assembly of under frames) rather than at original upstream entry points This modified SM is illustrated in the Simul8 screen shot of Figure 14 as SM1’ When compared to SM1, the number of ‘Job entry points’ is reduced from to 5, because of the DL family members share a case manufacturing system for which members of the consolidated group enter via a single entry point The reconfiguration of workflows through P/R couples illustrated by Figure 14 could be highly beneficial in BGF This is because the original BGF assembly system is overly flexible because set ups are only required for five rather than eight table types This example illustrates FFMS requirements and how a new conceptual system design can satisfy those focussed requirements In the developed virtual environment the above assumptions were confirmed when both SM1 and SM1’ simulations were run for month, with the same order entry rate The experiments performed with SM1 and SM1’ to test the above assumptions required a choice of model variables that suitably reflect BGF reality and allow a fair comparison to be made about the behaviours of the two model configurations Having studied historical BGF order patterns, an average inter-arrival time for products was calculated as being every 360 minutes To ensure parity the inter-arrival time for consolidated groups and was also maintained at the value but a batch of and respectively was input for those groups Conceptual design of improved simulation model SM1’ 79 International Journal of Computer Integrated Manufacturing Sample results related to this comparison are shown in the ‘lead time’ comparison presented in Table Significant improvement is predicted from the SM1’ configuration Further investigation of the source of this improvement was conducted Here, the source of the reduction in lead time was observed to be due to a significant reduction on queue (work in process) time in areas of the SM1’ system After-running the model with the aforementioned input conditions, screen shots of SM1 and SM1’ are shown in Figure 15 and Figure 16 It is evident that the Figure 14 Table major lead time difference comes from very significant WIPs and Queues in SM1 Figure 15 shows these WIPs (or queues) before the ‘Bench1’ work centre and the ‘CNC router’ work centre For the case of SM’, Figure 16 shows a reduced bottle neck in front of the ‘Sand top’ work centre To compare the total average queuing time (which can be a measure of WIP queuing time) for both SM1 and SM1’ systems, these three significant bottleneck work centres were selected for further analysis An associated summary of queuing times is illustrated in Table Simulation developed model of Drop Leaf Table assembly process SM1’ Lead time consumption comparison of SM1 and SM1’ Lead time in SM1 (days) Lead time in SM1’ (days) DL1 DL2 DL3 DL4 DL5 DL6 DL7 DL8 27.1 7.7 27.2 7.7 28.2 7.2 27.2 7.8 27.6 7.6 27.5 7.6 27.6 7.6 27.4 7.5 Figure 15 After run screen shot of SM1 (bottle neck identification) Figure 16 After run screen shot of SM1’ (bottle neck identification) 80 Z Cui and R Weston The same methodology was used to understand and redesign the BGF assembly systems for FH table types As explained earlier, there are seven different types of FH table produced by BGF This led to the creation of a case EoSS simulation model, namely SM2 SM2 therefore was created with seven FH table type entry points to create a simplified model of the BGF FH table class assembly Subsequent analysis showed that very minor processing route and time differences are required to assemble the FH table types Following detailed analysis of their respective processing routes and times, it was decided all seven FH table types could be reasonably consolidated into one product group This allowed the creation and simplified use of SM2’, a case system A SM2’ model screen shot is shown in Figure 17 This product type consolidation and resultant reconfiguration change also produced lead time spending benefits The results obtained from SM2 and SM2’ experiments were described previously in Cui and Weston (2008) Having created SM1’ and SM2’ the product consolidation ideas were further developed to consider the design of a case system capable of using a common system of resources to assemble all DL and FH table types Table Queuing time comparison of SM1 and SM1’ Categories Bench1 Q time (days) CNC router Q time (days) Sand top Q time (days) Sum of delay (days) Figure 17 SM1 SM1’ Change in queuing time 9.7 0.1 Reduced by 9.6 days 21.1 2.1 Reduced by 19 days 0.4 5.0 Increased by 4.6 days 31.2 7.2 Reduced by 24 days 5.3 Model reconfiguration After understanding BGF assembly system behaviours in response to different work dynamic impacts arising from DL and FH product families, the next stage of modelling carried out was to re-model the manufacturing system in a more realistic ‘as is’ form Because in the ‘real plant’ of BGF, all table assembly work is carried out in the ‘Table Assembly Area’, this means that in reality all types of table share the same Process and Resource system However, that system was understood to comprise an ‘as is’ configuration of more elemental P/R couples Therefore the actual situation was remodelled by creating SM3; such that both DL and FH product families share one assembly process, with the same set of work centres that are resourced by common underlying human and technical systems The screen shot of the developed SM3 is shown in Figure 18 The actual manufacturing system, modelled by SM3, realises products that cross product family boundaries Therefore during model runtime, instantaneous states of VD, PV and PMD all significantly impact on process routings and process timings Therefore resource requirements will vary as a function of time This was assumed to mean that reconfiguration of the manufacturing system at runtime is needed to produce all of those different types of table products in order to meet required due dates As can be seen in the screen shot of SM3 (see Figure 18) there are alternative routings out from the operation ‘collect components’, leading to different DL and FH table assembly processing Also, as the operation times for DL and FH tables are distinct, in SM3 it was necessary to distinguish the processes of the different product flows by using numeric labels Therefore when production orders arrive that define a requirement to Simulation model of Farm House Table assembly process SM2 International Journal of Computer Integrated Manufacturing Figure 18 81 ‘As is’ model of table assembly system SM3 assemble different types of DL and FH table, this means that a greater scope of PMD is required than was the case for SM1 and SM2; and that essentially a run time reconfiguration of ‘P/R couples’ is needed Therefore, the ‘as is’ manufacturing system is considered to be a case system; and SM3 which mimics the reality of the ‘as is’ situation is a case system model The foregoing research and discussion about manufacturing industry needing more rapidly to respond to changing business environments explained that Leagile and Postponement strategies (Mason et al 2000, Loe 1998) are becoming widely adopted in different types of ME In the case of BGF, their realisation of EoSS is necessary to achieve cost effectiveness in their business environment where many different products (circa 350 in the real situation) are required in uncertain quantities in any given time frame Also, BGF are heavily constrained as they have a limited human resource with the competencies required to produce such a wide product scope This paper has shown how that complex reality can be decomposed into separate sub-realities associated with different product types, and product families so that behaviours of different product realisations can be much better understood Following which (also using simulation technologies), those separate realities can be reconstituted back to mimic (and hence understand and analyse) situations that are closer to the reality The current paper has discussed alternative designs of BGF’s table assembly shop via implementing state of the art manufacturing concepts, as illustrated in Figure 19 To be able to design an effective manufacturing system that can realise economies of scope, it is necessary to adopt suitable manufacturing technologies so that the productivity can be enhanced Hence simulation models of possible ‘to be’ manufacturing system configurations can allow industry to ‘try out’ their assumptions about implementing different manufacturing paradigms to support medium and long term decision making about the design of manufacturing systems Mass customisation (Tseng 2001) is one of a number of possible manufacturing strategies that can be used to re-design the BGF manufacturing system When VD; PV and PMD impacts arise with the input of production orders, this can lead to unacceptably long lead-times The idea of Mass customisation is to realise Mass production for as much of the product realisation process as possible, so as to reduce the set up times of machines and human resources, to minimise queuing times and travelling time of work items, etc Following this, the idea is to postpone customisation to as late as possible (Jiao et al 2004) Therefore, to facilitate such a redesign at BGF a ‘to be’ model SM3’ was created to predict possible outcomes from using mass customisation in the table assembly shop in BGF A screen shot of SM3’ is shown in Figure 20 From the earlier SM modelling studies, it was observed that the ‘sand leg’ operation is needed for both Drop Leaf and Farm House tables Also observed was that the operations: ‘build frame’; ‘fit under frame to top’ and ‘assemble under frame’; are more customised operations This is because they are different operations for the different types of DL and FH tables So that in SM3’, the work centre used to 82 Z Cui and R Weston Figure 19 Illustration of the use of manufacturing concepts to achieve EoSS Figure 20 ‘To be’ model of table assembly system SM3’ ‘Sand legs’ was placed in the upstream process rather than in the downstream process 5.4 Related KPI comparison SM3 and SM3’ were both run over a period of one month using the same variations in production order data Following this, their relative behaviours and performances were observed Here primary KPIs were chosen systematically to differentiate between the manufacturing system designs coded by those SMs Simulated production behaviours led to runtime calculations related to these KPIs so that simulation model users can compare: 1) Lead time spending in 83 International Journal of Computer Integrated Manufacturing each model; 2) Utilisations of machines and human resources; 3) Throughput During experimentation it was observed that lead-time spending was important in terms of benchmarking the pros and cons of different configurations Illustrative results obtained when comparing SM3 and SM3’ are summarised in Table The order pattern input to SM3 and SM3’ to generate Table was as follows: DL and FH table orders come everyday with a batch of and respectively The total lead-time behaviours of SM3 and SM3’ are summarised by Table This illustrates example rules that can be applied to decide what assembly configuration would be best to adopt for a given particular group of customer orders or an expected new order pattern If for example mostly DL2, 3, 5, 6, or FH3, 5, product types are required by customers within a specified timeframe, then the SM3’ configuration will perform better But when product types DL1, 4, or FH 1, 2, 4, are dominant, the SM3 configuration will be a more suitable choice from a lead-time point of view Clearly however, the practicality of applying these rules will depend on the ease and extent to which reconfiguration and/or reprogramming is necessary and can be achieved Also, trying to reduce the lead-time may not be the best policy for MEs operating in environments where response times are not at a premium Nonetheless, in general this kind of knowledge about product/work dynamic impacts can assist production planners and production system designers, who might decide to switch configurations regularly or episodically by weighing up performance benefits relative to the cost of such changes Throughput was also chosen as a KPI to compare the behaviours and performance of different configurations Here benchmarking was carried out in which throughputs of SM1, SM2, SM3 and SM3’ were Table (1) Low order rate and Low batch size This was characterised as FH table orders come every days with a batch size of 5; while DL table orders come every days with a batch size of (2) Low order rate, High batch size This was characterised as FH table orders come every days with a batch size of 20; while DL table orders come every days, with a batch size of 15 (3) High order rate, Low batch size Characterised by FH table orders come every 0.5 days with a batch size of 5; DL table orders come every day with a batch size of (4) High order rate, High batch size Characterised by FH table orders come every 0.5 days with a batch size of 20; while DL table orders come every day with a batch size of 15 Figure 21 shows that with order pattern 1, namely when both frequency of orders and volume of products is low, SM3 and SM3’ configurations generate similar throughputs But when the order patterns changed the throughput behaviours of SM3 and SM3’ systems will react significantly differently To summarise, when the orders for FH and DL table types are high the SM3’ configuration is significantly better in terms of throughput and thence revenue generation But with lower order rates the SM3 configuration may perform better Investigation of this phenomenon showed that for the SM3’ configuration the ‘assembly capacity’ that can be realised with a fixed human and machine resource is greater than the other configurations Hence the case 3, proposed ‘to be’ assembly system (SM3’) could be used Summary of Lead time comparison between SM3 and SM3’ LT in SM3 (days) LT in SM3’ (days) Improved in LT Table compared All models were run with the same four scenario conditions The four different conditions were: DL1 DL2 DL3 DL4 DL5 DL6 DL7 DL8 FH1 FH2 FH3 FH4 FH5 FH6 FH7 15.1 15.3 14.8 14.4 Ö 15.0 14.6 Ö 12.9 14.5 27.3 17.6 Ö 15.2 14.1 Ö 15.4 14.5 Ö 9.5 14.5 14.1 16.1 12.6 15.8 14.8 14.4 Ö 13.8 15.8 13.9 6.9 Ö 15.1 14.8 Ö 14.2 14.7 Total LT behaviours observation in both SM3 and SM3’ LT in SM3 – LT in SM3’ (days) DL2,3,5,6,7 DL1,4,8 FH3,5,6 FH 1,2,4,7 12.5 76.8 7.6 77.7 84 Figure 21 Z Cui and R Weston Throughput comparison of four models with advantage by BGF, particularly if it has opportunities to grow The cost comparison has been made by including the cost of: human labour and machine depreciation; and materials But no account has been taken of factory overheads or work in progress costs This is because all relevant factory data was not available The assumptions made have not negated the argument but rather this has helped retain clarity of argument The message of this paper should be made clear Namely, by using the Model-Driven Method proposed, in theory much improved understanding of complex manufacturing realities can be achieved Potentially, this can lead to enhanced manufacturing systems design and engineering decisions; based on quantified performance criteria It is important to note that there is no constant best solution, the design or redesign of the manufacturing system is strictly connected to nature of the product dynamics in a given ME International Journal of Computer Integrated Manufacturing Conclusions and future research This paper has contributed to the literature in its field by: (1) Describing new thinking about relationships between customer related dynamics and how such factors can be expected to influence: (a) the design of EoSS manufacturing systems (b) the behaviours that can be generated by such systems and (c) needs to program and configure them (2) Introducing a systematic approach to decomposing and representing EoSS manufacturing systems This illustrates the use of de-coupling ideas that relate ‘process’, ‘resource’ and ‘work’ sub-systems It also illustrates how such subsystems can be modelled from different and relevant points of view, so as to provide explicitly defined, yet flexible, process-oriented couplings between models of product or work dynamics and modelled designs of (human and machine) resource systems, and their programming and configuration (3) A new Model-Driven Method of conceptually designing EoSS manufacturing systems is proposed which promises wide potential industrial application as a means of quantitatively predicting (i) relative performances of alternative manufacturing system designs and (ii) the likely cost of needed manufacturing system change associated with product/ work dynamics (4) A report is made on the initial case study testing of the concepts and Model-Driven Method introduced by this paper in which problems and data have been drawn from a home furniture producing ME The future work planned is described in outline in section 4.2 of this paper Currently, the authors are conducting further case testing in two significantly different MEs One of these MEs is located in Southern China, is a major producer of industrial air conditioning systems and will be referred to as AirCon International AirCon International manufactures many different product types (over 2000 types) which correspond to so-called ‘standard’, ‘customised’ and ‘special’ product groupings, where each group is subject to very significant volume dynamics One key innovation being tested in AirCon which extends use of the Model Drive Method reported in this paper concerns the use of new ‘Role’ and ‘Dynamic Producer Unit (DPU)’ modelling concepts (Weston et al 2009) 85 These new modelling ideas were conceived to explicitly represent configurations of process and resource subsystems that can be reused as virtual models of EoSS manufacturing systems These concepts have been used (i) to represent and decompose operational processes, used by AirCon to make their many product types and (ii) conceptually design new improved assembly manufacturing systems, where those designs are explicitly represented and structured as configurations of reusable DPUs Major benefits are being realised for AirCon by using EoSS manufacturing system models to support: the selection and implementation of new manufacturing paradigms; and the design of model driven integrated planning and procurement systems Also, key design inputs are being made in support of AirCon doubling its size and physically moving its manufacturing facilities The AirCon case is to be described in a sister refereed journal paper which builds on ideas introduced by this paper Acknowledgement The authors wish to thank their colleagues in MSI for the many discussions held and the data they have provided Particular thanks are due to Dr A Rahimifard for her help with related manufacturing paradigm knowledge, new modelling ideas related to DPUs and general simulation modelling help Thanks are also due to K A Kodua for his insightful and extensive inputs to the Air Con project References AMICE, 1993 CIMOSA: Open system architecture for CIM 2nd ed Berlin: Springer-Verlag Angra, S., Sehgal, R., and Noori, Z.S., 2008 Cellular manufacturing – A time-based analysis to the layout problem International Journal of Production Economics, 112 (1), 427–438 Apple, J.M., 1977 Plant Layout and material handling New York: John Wiley and Sons Azab, A and ElMaraghy, H.A., 2007 Mathematical Modeling for reconfigurable process planning Cirp Annals-Manufacturing Technology, 56 (1), 467–472 Azab, A and ElMaraghy, H., 2007 Sequential process planning: A hybrid optimal macro-level approach Journal of Manufacturing Systems, 26 (3–4), 147–160 Burns, J.R., 2001 Simplified translation of CLDs into SFDs Proceedings of the 19th Int Conference of the System Dynamics Society, Atlanta, GA Cui, Z and Weston, R.H., 2008 Model driven engineering of economy of scope systems Advanced Design and Manufacture to Gain a Competitive Edge – New Manufacturing Techniques and their Role in Improving Enterprise Performance, 697–706 Du, X., Jiao, J., and Tseng, M.M., 2003 Identifying customer need patterns for customization and personalization Integrated Manufacturing Systems, 14 (5), 387– 396 ElMaraghy, H.A., 2007 Reconfigurable process plans for responsive manufacturing systems Digital Enterprise Technology: Perspectives and Future Challenges, 35–44 86 Z Cui and R Weston ElMaraghy, H.A., 2005 Flexible and reconfigurable manufacturing systems paradigms International Journal of Flexible Manufacturing Systems, 17 (4), 261–276 ElMaraghy, H.A., 2009 Changing and evolving products and systems – models and enablers In: Hoda A ElMaraghy, ed Changeable and Reconfigurable Manufacturing Systems London: Springer-Verlag, 25–46 Fan, I.S and Liu, C.L., 1999 Product family and variants: Definition and models In: J Ashayeri, W.G Sullivan, and M.M Ahmad, eds Flexible Automation and Intelligent Manufacturing New york: Begell house, 213–224 Forrester, J.W., 1961 Industrial dynamics Cambridge: MIT Press Galan, R., et al., 2007 A systematic approach for product families formation in reconfigurable manufacturing systems Robotics and Computer-Integrated Manufacturing, 23 (5), 489–502 He, D., Kusiak, A., and Tseng, T.L., 1998 Delayed product differentiation: A design and manufacturing perspective Computer-Aided Design, 30 (2), 105–113 ISEE, 2007 ISEE Systems, USA Jiao, J.X., Zhang, L.F., and Prasanna, K., 2004 Process variety modelling for process configuration in mass customization: An approach based on object-oriented Petri nets with changeable structures International Journal of Flexible Manufacturing Systems, 16 (4), 335– 361 Koren, Y., 1983 Computer Control of Manufacturing Systems New York, NY, USA: McGraw Hill, Inc Koren, Y., et al., 1999 Reconfigurable manufacturing systems Cirp Annals 1999: Manufacturing Technology, 48 (2), 527–540 Lee, H.L., 1998 Postponement for mass customization In: J Gattorna, ed Strategic Supply Chain Alignment Aldershot: Gower, 77–91 Laakko, T and Mantyla, M., 1994 Feature-based modelling of product families In: ASME International Computers in Engineering Conference, 1, 45–54 Mason, J.R., Naylor, J.B., and Towill, D.R., 2000 Engineering the Leagile Supply Chain International Journal of Agile Manufacturing Systems, (2), 54–61 Meyer, M.H and Lopez, L., 1995 Technology strategy in a software products company Journal of Production Innovation Management, 12 (4), 194–306 Meyer, M.H and Utterback, J.M., 1993 The product family and the dynamics of core capability Sloan Management Review, 34 (3), 29–47 Mehrabi, M.G., Ulsoy, A.G., and Koren, Y., 2000 Reconfigurable manufacturing systems: Key to future manufacturing Journal of Intelligent Manufacturing, 11 (4), 403–419 Mohanty, R.P and Wenkatraman, S., 2003 Justification study for computer integrated manufacturing International Journal of Computer Integrated Manufacturing Systems, 16 (6), 366–374 Moodie, C., Uzsoy, R., and Yin, Y., 1995 Manufacturing cells: A systems engineering view London: Taylor and Francis Pierre de lit and Alain Delchambre, 2003 Integrated design of a product family and its assembly system Boston: Kluwer Academic Publishers Rahimifard, A and Weston, R.H., 2007 The enhanced use of enterprise and simulation modelling techniques to support factory changeability International Journal of Computer Integrated Manufacturing, 20 (4), 307–328 Simul8 Corporation, 2000 Simul8 User’s Manual Simul8: Herndon, VA Suresh, N.C and Kay, J.M., 1998 Group Technology and Cellular Manufacturing: Updated Perspectives In: N.C Suresh and J.M Kay, eds Group technology and cellular manufacturing – A state-of-the-art synthesis of research and practice Boston: Kluwer Academic Publisher, 1–14 Terkaj, W., Tolio, T., and Valente, A., 2009a Focused flexibility in production systems In: Hoda A ElMaraghy, ed Changeable and Reconfigurable Manufacturing Systems London: Springer-Verlag, 47–70 Terkaj, W., Tolio, T., and Valente, A., 2009b Design of Focused Flexibility Manufacturing Systems (FFMSs) In: Tullio Tolio, ed Design of Flexible Production Systems Berlin: Springer-Verlag, 137–190 Tseng, M.M and Jiao, J., 2001 Mass Customisation In: G Salvendy, ed 3rd ed Handbook of Industrial Engineering: Technology and Operations Management New York: Wiley, 684–709 Vernadat, F.B., 1996 Enterprise modeling and integration: Principles and applications London: Chapman and Hall Vernadat, F.B., 1993 CIMOSA: enterprise modelling and enterprise integration using a process-based approach In: H Yoshikawa and J Goossenaerts, eds Information infrastructure systems for manufacturing North-Holland, Amsterdam: Chapman and Hall, 65–84 Wang, B., et al., 1997 Integrated product, process and enterprise design: why, what and how? In: Ben wang, ed Integrated Product, Process and Enterprise Design London: Chapman and Hall, 1–20 Watkins, H.S., 1997 Developing customer-focused new product concepts In: Ben Wang, ed Integrated Product, Process and Enterprise Design London: Chapman and Hall, 21–44 Weston, R.H., Byer, N., and Ajaefobi, J.O., 2003 EM in support of team system engineering Concurrent Engineering: Enhanced Interoperable Systems, 865–872 Weston, R.H., et al., 2009 On modelling reusable components of change-capable manufacturing systems Proceedings of the Institution of Mechanical Engineers Part B-Journal of Engineering Manufacture, 223 (3), 313– 336 Weston, R.H., 1999 Model driven, component based approach to re-configuring manufacturing software systems International Journal of Operations and Production Management, 19 (8), 834–855 Weston, R.H., et al., 2007 Simulating dynamic behaviours in manufacturing organisations In: IESM 2007 International Conference on Industrial Engineering and Systems Management, Beijing, China [...]... Esper, T.L and Williams, L.R ., 2003 The value of Collaborative Transportation Management (CTM): its relationship to CPFR and information technology Transportation Journal, 42 (4 ), 55–65 Holweg, M ., Disney, S ., Holmstro¨m, J ., and Sma˚ros, J ., 2005 Supply chain collaboration: making sense of the strategy continuum European Management Journal, 23 (2 ), 170– 181 Hung, M.H ., Cheng, F.T ., and Yeh, S.C ., 2005... number of jobs; ni number of operations of job i; m type number of various resources; rs number of resources of type s, s 2 [ 1,2 , ,m]; Ri set of pairs of operations {k,l} belonging to job i, operation k precedes l; Qi set of pairs of operations {k,l} belonging to job i; Nq set of operations requiring resource q, q 2 [ 1,2 , r]; H large enough positive number; til processing time of operation l of job... C ., and Pramatari, K ., 2005 Using Web-Services to support collaboration practices along the supply chain In: Proceeding of the International Conference on Next Generation Web Services Practices Shen, D ., Yu, G ., Kou, Y ., Nie, T ., and Zhao, Z ., 2007 Resolving heterogeneity of Web-service composition in network manufacturing based on ontology International Journal of Computer Integrated Manufacturing, ... called operation k in short if no confusion is caused;  ykl ¼ 1 if operation k precedes operation l 0 otherwise where {k,l} 2 Qi, i ¼ 1,2 , ,n;  zij ¼ 1 if operation i precedes operation j 0 otherwise where {i,j} 2 Nq, q ¼ 1,2 , , rs, s ¼ 1,2 , , m Note: i 2 [ 1, , n]; s 2 [ 1, , m]; free operation means operation without precedence restrictions from technological planning ni n X X xil À ni n... Management, 23 (7 ), 794–818 Meixell, M.J and Gargeya, V.B ., 2005 Global supply chain design: A literature and critique Transportation Research Part E: Logistics and Transportation Review, 41 (6 ), 531–550 Mentzer, J.T ., Foggin, J ., and Golicic, S ., 2000 Collaboration: the enablers, impediments, and benefits Supply Chain Management Review, September/October, 52–58 Michalakos, S ., Anagnostopoulos, A ., Depasta,... http://www.w3.org/TR/ws-arch/#whatis [Accessed May 2008] Yang, S.J.H ., Lan, B.C.W ., and Chung, J.Y ., 2005 A trustworthy Web services framework for business processes integration In: Proceedings of 10th IEEE International Workshop on Object-Oriented Real-Time Dependable Systems WORDS 200 5, pp 186–193 International Journal of Computer Integrated Manufacturing Vol 2 3, No 1, January 201 0, 20–39 A hybrid particle swarm optimisation... February 2006] Kalogeras, A.P ., Gialelis, J.V ., Alexakos, C.E ., Georgoudakis, M.J ., and Koubias, S.A ., 2006 Vertical integration of enterprise industrial systems utilizing Web Services IEEE Transactions on Industrial Informatics, 2 (2 ), 120–128 MacCarthy, B and Atthirawong, W ., 2003 Factors Affecting Location Decisions in International Operations – A Delphi Study International Journal of Operations and Production... positive number; til processing time of operation l of job i, l 2 [ 1, , ni]; xik starting time of operation k of job i, k 2 [ 1, , ni]; xsi starting time of the first (or free) operation of job i; xie completion time of the last (or free) operation of job i; ai availability time of job i; di delivery due date of job i; [i,k] the kth operation of job i, also called operation k in short if no confusion is... integration in holonic manufacturing systems Fuqing Zhaoa *, Yi Honga, Dongmei Yua, Yahong Yangb and Qiuyu Zhanga a School of Computer and Communication, Lanzhou University of Technology, Lanzhou 73005 0, Gansu, P.R China; bCollege of Civil Engineering, Lanzhou University of Techchnology, Lanzhou 73005 0, Gansu, P.R China (Received 29 January 2009; final version received 25 July 2009) Modern manufacturing systems... of Computer Integrated Manufacturing, 20 (2 ), 222–233 Simatupang, T.M and Sridharan, R ., 2004 Benchmarking supply chain collaboration: An empirical study Benchmarking: An International Journal, 11 (5 ), 484–503 Tyan, J.C ., Wang, F.K ., and Du, T.C ., 2003 An evaluation of freight consolidation policies in global third party logistics Omega, 31 (1 ), 55–62 W3C, 2004 Web Service Architecture Requirement