Tele Manu Fatur a 72

Int. J. Production Economics 62 (1999) 87}105

Agile manufacturing: A framework for research and development

A. Gunasekaran*

Department of Management, University of Massachusetts, North Dartmouth, MA 02747, USA

Abstract

Agile manufacturing (AM) is a new concept in manufacturing intended to improve the competitiveness of "rms.Manufacturing processes based on AM are characterized by customer}supplier integrated process for product design,manufacturing, marketing, and support services. This needs decision-making at functional knowledge levels, stable unitcosts, #exible manufacturing, easy access to integrated data, and modular production facilities. Agile manufacturingrequires enriching of the customer, co-operating with competitors, organizing to manage change, uncertainty andcomplexity, and leveraging people and information. In the recent years, a number of research papers have been publishedin the area of AM. However, a framework for the development of AM has not received due attention from bothresearchers and practitioners. Realizing the importance of agile manufacturing in the 21st century manufacturingcompetitiveness, an attempt has been made in this paper to review the literature available on AM with the objective to: (i)identify key strategies and techniques of AM, (ii) suggest some future research directions and (iii) develop a framework forthe development of agile manufacturing systems (AMSs) along four key dimensions which include strategies, technolo-gies, systems and people. ( 1999 Elsevier Science B.V. All rights reserved.

Keywords: Agile manufacturing; Review; Future research; Development; Framework

1. Introduction

Businesses are restructuring and re-engineeringthemselves in response to the challenges and de-mands of 21st century. The 21st century businesseswill have to overcome the challenges of demandingcustomers seeking high quality, low cost products,responsive to their speci"c and rapidly changing

*Corresponding author. Tel.: #1 508 999-9187; fax: #1 508999-8776; e-mail: [email protected].

needs [1]. Agility addresses new ways of runningcompanies to meet these challenges. Agility isabout casting o! those old ways of doing thingsthat are no longer appropriate } changing patternof traditional operation. In a changing competitiveenvironment, there is a need to develop organiza-tions and facilities signi"cantly more #exible andresponsive than current existing ones [2,3].

Agile manufacturing can be de"ned as the capa-bility of surviving and prospering in a competitiveenvironment of continuous and unpredictablechange by reacting quickly and e!ectively to chang-ing markets, driven by customer-designed pro-ducts and services [4]. Agile manufacturing is not

0925-5273/99/$ - see front matter ( 1999 Elsevier Science B.V. All rights reserved.PII: S 0 9 2 5 - 5 2 7 3 ( 9 8 ) 0 0 2 2 2 - 9

about small-scale continuous improvements, butan entirely di!erent way of doing business [5].Agile manufacturing is a new expression that isused to represent the ability of a producer of goodsand services to thrive in the face of continuouschange. These changes can occur in markets, intechnologies, in business relationships and in allfacets of the business enterprise [6]. Agile manu-facturing requires to meet the changing marketrequirements by suitable alliances based oncore-competencies, organizing to manage changeand uncertainty, and leveraging people andinformation.

Agile manufacturing is a vision of manufacturingthat is a natural development from the originalconcept of &lean manufacturing'. In lean manufac-turing, the emphasis is on cost-cutting. The require-ment for organizations and facilities to becomemore #exible and responsive to customers led to theconcept of the &agile' manufacturing as a di!erenti-ation from the &lean' organization. This require-ment for manufacturing to be able to respond tounique demands moves the balance back to thesituation prior to the introduction of lean produc-tion, where manufacturing had to respond to what-ever pressures were imposed on it, with the risks tocost and quality. The move to lean production fromagile and vice versa is a major challenging task[7,8].

Most of the literature tends to focus on a fewenablers/strategies of AM without a comprehensiveframework for achieving AM. In most cases, thoseagile strategies and techniques are disconnected.Such a comprehensive framework can be developedeither by "eld studies or by a systematic literaturesurvey to identify the key strategies and techniquesof AM and then integrate them to develop an AMS.Since AM is at the developmental phase, it is worth-while to develop the idea based on existing studies.Therefore, literature survey is the adopted meth-odology in this paper for the development ofa framework for an agile manufacturing system(AMS).

The organization of the paper is as follows:Section 2 presents the classi"cation of the literatureavailable and a brief review of the previous researchon AM. Comments on the literature and futureresearch directions are discussed in Section 3.

A framework for the design of AM is presented inSection 4. Section 5 concludes the paper.

2. Classi5cation and a review of previous researchon agile manufacturing

In this section, a classi"cation of the literatureavailable on AM and a brief review of each articleare presented. Agile manufacturing includes rapidproduct realization, highly #exible manufacturing,and distributed enterprise integration. Technologyalone does not make an agile enterprise. Every com-pany must "nd the right combination of culture,business practices, and technology that are necessaryto make itself agile. In this paper, the focus is on theoperating conditions of factories organized as #ex-ible networks of processors based on core competen-cies. With this scope in mind, an attempt has beenmade in this section to review the literature on AM.

2.1. Classixcation of the literature on agilemanufacturing

The literature available on AM has been classi-"ed based on the nature and the focus of agileenablers which include criteria such as strategies,technologies, systems and people. A classi"cationscheme based on the nature and application of themodels is proposed for easy understanding of theresearch work on AM. The literature available un-der each criterion have been further grouped undersub-classi"cation to improve to the clarity of thepresentation and highlight some key factors of AMunder each classi"cation. The main objective of thisparticular classi"cation is to develop a suitableframework for AMSs along these four dimen-sions/criteria. Agility should be in all areas ofmanufacturing to e!ectively respond to changingmarket requirements. Achieving agility thereforerequires #exibility and responsiveness in strategies,technologies, people and systems. Table 1 showsthe classi"cation of the literature on AM and thecorresponding references on the basis of strategies,technologies, systems and people. It is imperativeto mention that the literature surveyed in this paperare not exhaustive, but only a representative.

88 A. Gunasekaran/Int. J. Production Economics 62 (1999) 87}105

Table 1Classi"cation of agile manufacturing literature

Criteria for classi"cation of the literature Sub-classi"cation References

Virtual enterprise [9}17]Strategies Supply chain [1,5,7}9,18}22]

Concurrent engineering [7,17,20,23}26]

Technologies Hardware } tools and equipments [19,24,27}35]Information technologies [4,17,19,31,33,34,36}40]

Design systems [25,27,41}50]Systems Production planning and control systems [15,16,42,49,51]

System integration and database management [1,15,17,24,36,37,39,46,51}56]

Knowledge workers [17,56,57]People Top management support and employee empowerment [17,21,58,59]

Training and education [9,17,44,56,57,59,60]

2.2. Review of literature on agile manufacturing

In this section, the literature available on AMhas been reviewed based on the classi"cationscheme shown in Table 1. The stated articles havebeen reviewed for the key strategies adopted, tech-nologies used, organizational issues and humanresource development in AM. For the purpose ofbrevity, the review of each article has been kept tohighlighting the key strategies and techniques sug-gested by authors.

2.2.1. StrategiesStrategic approach to performance improvement

is gaining attention in all areas of manufacturing.The reason for this is that it takes into account thelong-term interest of the company in determiningsuitable business and operational policies. Agilemanufacturing itself is a strategy. To achieve this,several sub-strategies are needed including virtualenterprise, rapid-partnership formation, rapid pro-totyping, and temporary alliances based on corecompetencies. Without suitable strategies, tech-nologies and systems alone not su$cient to achieveagility. Agile manufacturing requires customer in-tegrated multidisciplinary teams, supply chainpartners, #exible manufacturing, computer-integ-rated information systems, and modular produc-tion facilities [61]. In this section, some of the keystrategies of AM have been identi"ed for the review

of literature and further details are presented here-under.

2.2.1.1. Virtual enterprise. A virtual organizationis the integration of core competencies distributedamong a number carefully chosen but real organ-izations all with similar supply chain focusing onquick to market, cost reduction and quality [9].Generally, a single organization is often may not beable to respond quickly to changing market re-quirements. Temporary alliances or partnershipbased on core competencies of "rms will help toimprove the #exibility and responsiveness of organ-izations. However, co-ordination and integrationare complicated in this environment. Appropriatestrategies/methods such as communication, train-ing and education, and strategic alliances can beadopted for an e!ective co-ordination and integra-tion of participating "rms at di!erent levels of co-operation.

Virtual Manufacturing (VM) is an integrated,synthetic manufacturing environment exercised toenhance all levels of decision and control ina manufacturing enterprise. The agile enterpriserequires VM to respond to changing market re-quirements quickly. VM environments are beingproposed to improve their responsiveness, improveproduct and process design, reduce manufacturingrisks, improve manufacturing design and op-eration, support manufacturing system changes,

A. Gunasekaran/Int. J. Production Economics 62 (1999) 87}105 89

enhance product service and repair, increase manu-facturing understanding, and provide a vehicle formanufacturing training and research. The VirtualEnterprise (VE) environment places a number ofspecial requirements on the process design activity.Since virtual or distributed enterprises are tempor-ary, such organizations must be easily assembledand disassembled. Individual partner organizationsdo not cease to exist during their membership in theVE. This point highlights another important issue,security. These require appropriate industrial legis-lation and legal protection to be established [17].

Today, the manufacturing industry particularlythe One-of-a-Kind Production (OKP) industry,tends to be lean, agile and global. This tendencyleads to a new concept of a virtual company thatconsists of several sub-production units geographi-cally dispersed in the world as branches, joint ven-tures and sub-contractors. Many OKP companies,such as those in the heavy industry, for examplea shipbuilding company, have become virtual com-panies [16]. Another example is that constructionindustry has practised VEs for a number of years.The manufacturing also has become of more ser-vice oriented industries. Therefore, some of thesuccessful strategies and techniques used in con-struction such as for partnership formation canalso be implemented for the development of AM.

As noted earlier, partnership formation based oncore-competencies and temporary alliances facili-tate agility in manufacturing. In the partnershipformation, there is a need for information on threefunctions of AM that include prequalifying part-ners, evaluating a product design with respect tothe capabilities of potential partners, and selectingthe optimal set of partners for the manufacture ofa certain product. In particular, Herrmann et al.[12] present an information model that describesthe systems, process capabilities, and performanceof an AM "rm.

Agile manufacturing combines the strengths ofquali"ed partners in a virtual organization to meeta certain market need. High level process planningis a way to aid designers to improve their designwith respect to the capabilities of potential part-ners. An automated high level process planningsystem for mechanical components has been de-veloped by Gupta et al. [10] that considers both

global and partner-speci"c manufacturing capabili-ties so that cost, cycle time, and quality attributescan be derived for a given design. The system gener-ates a set of manufacturing operations and candi-date partner plants that can contribute to themanufacture of the candidate design. The mostpromising process plans resulting from this processmay be used to select the most suitable partners forthe VE. The article reviewed under VE highlightsthe communication and performance measurementas the key issues for developing an e!ective AMS.

2.2.1.2. Supply chain. A supply chain is the globalnetwork used to deliver products and services fromraw materials to end customers through an engin-eered #ow of information and physical distribution.The supply chain management system focuseson resolving business process problems that areimportant to the customers.

As producers, wholesalers and retailers seekmore e!ective ways of marketing their products,they increasingly examine their supply chains forways to reduce costs. The logistics supply chainaims to achieve improved #exibility by reducedsupply cost, reduced stock holding costs, removalof stock rooms and increased selling space for re-tailers, control of inbound materials, integration offunctions from purchasing to sales, and increasedcontrol of the supply chain [21]. For agility insupply chain, top management involvement is vitalto e!ectively reengineering supply chain andlogistics. In agile supply chain environments,relationship with suppliers and interaction be-tween suppliers should be #exible in terms ofdelivering products/services and responsive.Appropriate frameworks should be established todevelop a management system for agile manufac-turing enterprise incorporating the information#ow and performance measurements.

The global supply chain management includesEnterprise resource planning (ERP). Thoughtwareand groupware with &legacy systems' focusing onMRP II fundamentals such as global sales andoperations planning can be used to analyze andoptimize an entire supply chain from purchasing/suppliers through AM using a streamlined logisticsnetwork and overcoming cultural, communica-tions, and cross-functional obstacles [22]. The


barriers of cultural, communication and cross-functional obstacles can be overcome in a VE bytraining and education, strategic alliances, ad-vanced information technologies to improve thecommunication and team working by empoweredemployees. To be truly agile, supply partners mustbe able to move even more quickly and with moree$cient utilization of the existing equipment, exist-ing facilities and even existing designs [18].

Responsibility-Based Manufacturing (RBM) isa new architecture of production system that fallsunder the umbrella of the AM paradigm. In mass-customization environments, RBM allows mostadjustments for process and product variety to takeplace dynamically and rapidly during productionwithout the need for a priori system recon"gura-tion. Active resources (mobile robots, intelligentpallets, etc.) take the responsibility for the produc-tion of individual parts/products, thus implemen-ting the unitary relation, individual customer toindividual producer. As di!erent products are pro-duced at the same time, to achieve a coherent andaugmented team performance, coordination ofactive resources is obtained by exploiting therelations which hold among the activities of theindividual product process plans [19].

Agility in supply chain can be achieved by integ-rating organizations, people, and technology intoa meaningful unit by deploying advanced informa-tion technologies and #exible organizational struc-tures to support highly skilled, knowledgeable, andmotivated people. Although the words agile andlean are sometimes used interchangeably in thiscontext, they are not the same. Lean implies highlycost-e$cient and productive, although it does notnecessarily imply being responsive. Agile, on theother hand, stresses the importance of being highlyresponsive to meet the &total needs' of the customer,while simultaneously striving to be lean. Thus, anagile manufacturer places a higher priority onresponsiveness than cost-e$ciency while a manu-facturer whose primary goal is to be lean, com-promises responsiveness over cost-e$ciencies [53].

Supply chain management in VE needs di!erentset of frameworks, strategies, techniques and per-formance measurement criteria. For example, therelationship with suppliers in lean organizations isbased on long-term focusing on cost reduction. But

in the case of AM, the relationship is temporarfocusing on responsiveness. Therefore, appropriatesupply chain management strategies/methods andperformance measurements should be establishedto improve the e!ectiveness of supply chain man-agement in AM enterprises.

2.2.1.3. Concurrent engineering. Agility in manu-facturing requires a change around the formationof product development teams. These teams haveincluded representatives with di!erent expertise,such as design, manufacturing, quality assurance,purchasing, marketing, "eld service and support.Change has also included relaxing policies thatinhibited design changes and providing greaterauthority and responsibility to members of designteams. Managing change in a manufacturing envi-ronment requires a more systematic method of con-currently designing both the product and thedownstream processes for production and support.This systematic approach is fundamentally knownas Concurrent Engineering (CE). Mehdat andRook [26] discuss the need to achieve a multidis-ciplinary team working environment as a prerequi-site for facilitating a CE strategy. They alsoexamine the role of enabling processes and tech-niques such as Computer Aided Design, ComputerAided Engineering and formal methods such asDesign for Manufacture and Assembly, to achievereduced product development cycles, whilst im-proving the quality of products.

Agile manufacturing demands a manufacturingsystem to be able to produce e$ciently a largevariety of products and be recon"gurable to ac-commodate changes in the product mix and productdesigns. The manufacturing system recon"gurabil-ity and product variety are critical in AM. Theconcept of agility has an impact on the design ofassemblies. To implement AM, methodologies ofdesign for AM are needed. Design for agile assem-bly is accomplished by considering operationalissues of assembly systems at the early productdesign stage [25].

Agile manufacturing requires an intelligent en-gineering design support system that can providerapid evaluation of engineering designs and designchanges. Often this process results in modi"edproducts that require adjustments and retooling of


the manufacturing processes that produce theproduct. Graham and Ragade [23] discuss the re-quirements for an intelligent CE design supportstation that will allow the design engineer to evalu-ate design modi"cations while concurrently exam-ining suitability for manufacturing.

The articles reviewed under strategies for AMdeal with issues such as VE and strategic alliances(partnership formation) based on core-competen-cies, responsive logistics, rapid product design, and#exible computer systems. However, the main focusis on the virtual enterprise, strategic alliances andenterprise integration with computerized informa-tion systems. Since the system should be proactiveto changes in the market requirements, there isa need for long-term policies. The major issues indeveloping AM are how the resources can berecon"gured/reused and procured to face the chal-lenges of market dynamism, technological advance-ments, infrastructure, government policies andlegislation. In most cases, the overall objective is tominimize the capital expenditures and the cost ofresources in meeting the changing market require-ments with the objective to enhance the marketshare or pro"t level. In AM enterprises, strategiesshould be formulated based on top down approachand they should be implemented using bottom-upapproach. Therefore, issues such as market types,strategic alliances, capital investment decisionsshould rely on the top management and the imple-mentation rests on the functional level managersand employees.

2.2.2. TechnologiesIn a global manufacturing environment, in-

formation technology plays a dominant role ofintegrating physically distributed manufacturing"rms. Critical to successfully accomplishing AMare a few enabling technologies that includerobotics, Automated Guided Vehicle Systems(AGVSs), Numerically Controlled (NC) machinetools, Computer-Aided Design (CAD)/Computer-Aided Manufacturing (CAM), rapid prototypingtools, Internet, World Wide Web (WWW),Electronic Data Interchange (EDI), Multimediaand Electronic Commerce [4]. For example, vir-tual enterprise relies on technologies such as VM,CAD/CAM, and Information Technology (IT)

(including multimedia, rapid prototyping, Internet,Electronic Commerce). Some of the key agile-enabled technologies include mobile robots, intelli-gent pallets, and #exible "xtures, and strategic,tactical and operational performance measures areto be considered in assessing the impact of alterna-tives with the objective to select the most suitabletechnologies. Adamides [19] reviews the techno-logy involved in RBM and discusses the qualitativeand quantitative performance characteristics of itscoordination mechanism. Agile technologicalrequirements are discussed under hardware thatinclude equipments and tools, and informationtechnologies which include computers and software.

2.2.2.1. Hardware } equipments and tools. Agilemanufacturing requires rapid changeover from theassembly of one product to the assembly of a di!er-ent product. This in turn needs a rapid hardwarechangeover by robots, #exible part feeders, modu-lar grippers, and modular assembly hardware. Thedivision of assembly, feeding, and unloading tasksbetween multiple robots should be examined withpriortization based upon assembly time. Rapidsoftware changeover will be facilitated by a real-time, object-oriented software environment utiliz-ing graphical simulations for o!-line softwaredevelopment. An innovative dual (Virtual Manu-facturing Enterprise) bus (VMEbus) controllerarchitecture permits an open software environmentwhile accommodating the closed nature of mostcommercial robot controllers. These agile featurespermit new products to be introduced with mini-mal downtime and system recon"guration. For in-stance, Quinn et al. [35] discussed a design for AMworkcells intended for light mechanical assembly ofproducts made from similar components (i.e., partsfamilies).

Agile manufacturing needs intelligent sensingand decision making systems capable of automati-cally performing many tasks traditionally executedby human beings. Visual inspection is one such taskand there is a need for e!ective automated visualinspection systems in AM environments [30]. Agilemanufacturing requires agile-enabling technologiessuch as virtual machine tools, #exible "xturing, andagile design alternatives [29]. As noted earlier,physically distributed manufacturing environments/


VEs demand high level communication systemssuch as Internet, EDI and Electronic Commerce toexchange information at various levels of manufac-turing organizations. Flexible "xturing is a keytechnology in the integration of AMS and the lackof e!ective #exible "xturing can be a signi"cantobstacle to implementation [27].

An agile workcell has been proposed for lightmechanical assembly. The workcell includes mul-tiple robots, a conveyor system, multiple #exibleparts feeders at each robot's workstation,Computer-Controlled Digital (CCD) cameras forparts feeding and hardware registration, and a dualVMEbus control system. A #exible parts feederdesign uses multiple conveyors to singulate theparts and machine vision to locate them. Special-ized hardware will be encapsulated on modulargrippers and modular worktables that can bequickly interchanged for assembly of di!erentproducts [33]. Towards achieving agility in manu-facturing, Lee [31] has discussed the recon"gura-bility of a manufacturing system. It is analyzedbased on the relationship of component routes,material handling costs, recon"guration cost, andso on. Components with similar routes are selectedin an early design stage in order to minimize thenumber of machines to be relocated. The variety ofresources required is reduced by a proper selectionof components and manufacturing processes forsystem recon"guration.

2.2.2.2. Information technologies. Information Tech-nologies such as Internet, CAD/CAM, MRP, ERP,EDI, EC can be employed for an e!ective integra-tion of physically distributed "rms in AM enter-prises. Flexible simulation software systems willsurely enhance the e!ectiveness of VEs in a phys-ically distributed manufacturing environment andhence agility in manufacturing. Therefore, VirtualReality Software (VRS) for robotics and manufac-turing cells simulation can be employed to obtaina 3D graphics model for manufacturing lines. A 3Dgraphics model will help facility planners to visual-ize the system before constructing it, make alter-native designs, program robot paths, obtainlayouts for the systems, obtain data for the discreteevent simulation, and develop the cell controlprogram [34]. As indicated earlier, virtual enter-

prises require high level communication systems in-cluding Internet, Multimedia and CAD/CAM toeliminate non-value adding activities in the supplychain. Also, this helps to avoid human related errorsin exchanging information and controlling variousproduction and operations in AM environments.Agile manufacturing demands knowledge workerssuch as computer operators, design engineers, soft-ware engineers, systems analysts and corporate plan-ner. Therefore, there is a need to identify the typeand level of skills required in AM environments.

Merat et al. [33] have proposed an agile workcellfor light mechanical assembly. Object-orientedsoftware (C##) running under VxWorks, a real-time operating system, can be used for workcellcontrol. In this way, agile software architecture wasdeveloped for rapid introduction of new assembliesthrough code re-use. Incorporated in the softwarearchitecture is a capability for simulation of theworkcell so that controller software could be writ-ten and tested o!-line, enabling the rapid introduc-tion of new products which facilitates agility inmanufacturing. Global access to people, data, soft-ware, documents, and multimedia has allowedorganizations to shorten the development cycle fornew products, communicate with experts from allover the world, improve manufacturing processes,and receive immediate customer feedback [32].

The need to rapidly recon"gure an AMS makesit especially di$cult to test the system's controlsoftware. Although thorough testing is essential forsystem reliability, the time available for testing maybe short. With a simulator, however, the softwarecan be developed and tested independently fromthe actual workcell, while production continues orthe workcell is recon"gured for the next targetproduct. To facilitate testing of AM software, a 3Dgraphical simulator can be used. It permits work-cell control software to be tested with a virtualworkcell, which exhibits much of the behavior ofthe real workcell. Considerable time and e!ort canbe saved by simulating various workcell scenarios,some of which are di$cult to create in a real work-cell, for example device failure [39].

Yang [38] proposed an object-oriented model ofan AMS with a de"nition of the agile objects atfour levels and their features. Meanwhile, it ex-plains the process in which the agile objects, under


the stimulation of tasks (market demands) get as-sembled into objects at higher levels and are integ-rated into agile system by sending information toeach other and by accepting information selective-ly. Here the object-oriented method is adopted toclarify that the agile system is a #at dynamic net-work in structure, and that in working mechanismsit mainly adapts to the quick market changes byrapid reengineering. The principles of AM can beused to demonstrate the feasibility of customizedspaceframe vehicle manufacture using robotic cel-lular manufacturing. The concepts used can bevalidated using computer-aided simulation anda full scale manufacturing cell. However, furtherworks need to be done in this respect. Areas ofinvestigation should include system design, joiningtechnologies, tooling design and manufacturingcontrol systems [40].

Bocks [36] developed a Data ManagementFramework (DMF) to support agility in manufac-turing. A DMF can be de"ned as the ability of anenterprise to manage and distributed data, in-formation, and knowledge as the decisive enablerfor core enterprise business process. The purpose ofDMF is to provide a seamless enterprise data man-agement solution in support of the AM environ-ments. Integration of current fragmented computersystems, causing over-complexity, is perhaps thebiggest challenge the AM enterprise faces.

Wang et al. [37] present an Internet assistedmanufacturing system for AM practice. This systemuses Internet as an interface between a user and theCentral Network Server (CNS) and allows a localuser to operate remote machines connected to theInternet It consists of a CAD/Computer-AidedProcess Planning (CAPP)/CAM/Computer AidedAnalysis (CAA) integrated CNS which links tolocal Flexible Manufacturing Systems (FMS), orComputer Numerically Controlled (CNC) machin-es by means of cable connections. After a local userinputs the product information, the CNS can gen-erate complete CAD/CAPP/CAM/CAA "les andcontrol the remote FMS or CNC machine toaccomplish the whole production process.

From the review of agile-enabling technologies,it can be easily noted that the selection of technolo-gies for achieving agility in manufacturing dependsupon the strategies that are selected to meet chang-

ing market requirements. For example, JIT mayrequire EDI and FMS needs AGVs, Robots, andNC machine tools. Nevertheless, agility heavilyrelies on virtual manufacturing/enterprise or phys-ically distributed manufacturing environments.Therefore, technologies such as IT, manufacturingcells, robots, #exible part feeders, modular assem-bly hardware, automated visual inspection system,virtual machine tools, #exible "xturing, CAD/CAM and automated high-level process planningare essential for developing AMSs.

2.2.3. SystemsThe systems for AM should include mostly soft-

ware/decision support systems for various planningand control operations including materials require-ments planning, design, manufacturing resourceplanning, scheduling, and production planning andcontrol. Based on the nature of AM environments,an attempt has been made in this section to discussthe various control systems required for AM envi-ronments. There are several computer-integratedsystems that could be used for AM, some of themare as follows: (i) MRPII, (ii) Internet, CAD/CAE,(iii) ERP, (iv) Multimedia, and (v) Electronic Com-merce.

2.2.3.1. Design systems. Agile manufacturing re-quires rapid product design systems with the objec-tive to switch over to new products as quickly aspossible. This in turn needs a system to groupvarious resources and products with the objectiveto reduce the non-value adding activities and hencethe time to reach markets with right products at theright time. Candadai et al. [43] discuss a variantapproach for quick design evaluation in an AMenvironment. Their approach uses a STEP-basedproduct model to generate the Group Technology(GT) code of a candidate product design, and addi-tional information critical to the product's manu-facture. This information is used to conduct ane$cient search for similar products manufacturedby potential partners and to obtain useful feedbackon manufacturing processes, production times,costs, and quality attributes of these products. Suchfeedback is valuable for design evaluation andimprovement early in the design cycle of a productin AM.


Monsplaisir [50] describes the evaluation of twoComputer Supported Cooperative Work (CSCW)prototypes to aid engineering teams in the design ofan AM facility. The CSCWs facilitated considera-tion of a large number of #exibility and agilitycriteria associated with the design of manufacturingsystems. Both prototypes support the functionssuch as anonymous inputs, parallel processing ofinformation, group memory, electronic brainstorm-ing, and consensus building. One of the two CSCWprototypes included three enhancements to furtherimprove consensus building in the design teams.The laboratory test results indicate that bothCSCW prototypes provided e!ective support fordesign teams.

As discussed earlier, both hardware and softwarerecon"gurability are required to achieve agility inmanufacturing. Kusiak and He [25] developedthree rules applicable to the design of products foragile assembly from an operational perspective.These rules are intended to support the design ofproducts to meet the requirements of AM. In orderto reduce time to markets, CE principles can beemployed in product development process in AMenterprise.

2.2.3.2. Production planning and control systems.For virtual companies, traditional production con-trol and management systems, methods and the-ories do not satisfy their needs for productionplanning and control. The following aspects are tobe considered for PPC in AM environments: (1)modelling of evolutionary and concurrent productdevelopment and production under a continuouscustomer's in#uence; (2) real-time monitoring andcontrol of the production progress in a virtualcompany; (3) a #exible or dynamic company con-trol structure to cope with uncertainties in themarket; (4) adaptive production scheduling struc-ture and algorithms to cope with uncertainties ofproduction state in virtual company; (5) modellingof production states and control system in a virtualcompany; and (6) the reference architecture fora virtual company [16].

The #exibility of an AMS can be achieved in partthrough computer software. The system's controlsoftware must be adaptable to new products and tonew system components without becoming unreli-

able or di$cult to maintain. This requires design-ing the software speci"cally to facilitate futurechanges. Kim et al. [49] have developed a softwarearchitecture for control of an AM workcell, anddemonstrated its #exibility with rapid changeoverand introduction of new products. AMSs can bene-"t signi"cantly from a database support for partnerselection, and mistake proo"ng.

Lee et al. [51] describe an AM Database System(AMDS) designed for capturing and manipulatingthe operational data of a manufacturing cell.AMDS is a continuous data gathering, real-timeData Base Management System (DBMS), and itcan be logged either locally or remotely and usedfor o!-line analysis as well. The temporal opera-tional data obtained can be used for performanceand reliability analysis, high-level summary reportgeneration, real-time monitoring and active inter-action.

Song and Nagi [15] propose a framework forproduction control in an AM environment inwhich: (1) information is modelled in a hierarchicalfashion using Object Oriented Methodology(OOM); (2) information transactions are speci"edby the work#ow hierarchy consisting of partnerwork#ows; (3) information #ow between partners iscontrolled by a set of distributed Work#ow Man-agers (WM) interacting with partner knowledgebases, which re#ect partner speci"c informationcontrol rules on internal data exchange, as well asinter-partner mutual protocols for joint partnercommunications; and (4) the prototype system isaccomplished using the World Wide Web based ona client}server architecture. Systems such as ERP,EDI and EC can be employed for operations con-trol activities of AM enterprises after making ap-propriate modi"cations to those systems with theobjective to incorporate the system speci"c charac-teristics of AM.

2.2.3.3. Data management and system integra-tion. Agility imposes special requirements on theinformation systems used to run an enterprise. Inaddition to satisfying the traditional requirements,an agile enterprise information system must be ableto be recon"gured in a very short time and shouldbe able to include parts of information systemsfrom other companies if a Virtual Corporation is


required to meet the market demand. The tradi-tional, common shared data base model of integra-tion has severe problems in this environment. TheSystems Integration Architecture (SIA) is based ona new, transformation model of integration andprovides sets of high level services which allowinformation system modules, including foreignmodules, to be rapidly recon"gured. These servicesinclude the management of information (naming,locations, format, "les structure and data accesstype) as well as the relationships among data sets,communications between objects on heterogeneouscomputer systems, wrappers for legacy software,diverse control approaches, composition of mod-ules into processes and user interfaces. SIA alsoshould allow for the integration of functions andcontrol as well as data [14].

Rapid software changeover can be facilitated bythe use of a real-time, object-oriented softwareenvironment, modular software, graphical simula-tions for o!-line software development, and aninnovative dual VME (Virtual ManufacturingEnterprise) bus controller architecture. These agilefeatures permit new products to be introduced withminimal downtime and system recon"guration[55]. System integration in AM is complicated be-cause of the nature of a virtual and physicallydistributed enterprise. To be true to its concept,AM needs to adopt the principles of all processadvancements evolving out of machine tool andcutting tool technology and their related scienti"c"elds [24].

McMullen [54] shows how the philosophies,practices, decision processes, measurements, logis-tics, and systems architectures of Theory of Con-straints (TOC) all work together to provide aninfrastructure for AM. It is suggests that TOCsystems can be moved to a co-standard status withtraditional MRP/Capacity Requirements Planning(CRP) systems, in order to encourage the systemscommunity to provide the MRP II and ERP sys-tems infrastructure required to support the emerg-ing agile manufacturers.

One of major issues that should be resolved,namely what sort of performance measures wouldbe suitable for selecting partners based on tempor-ary alliances and changing market needs. Sinceresponsiveness is the key objective, companies may

demand a set of simple performance measures indeveloping AMSs based on both the "nancial andnon-"nancial performance measures. CE can beused for the optimal selection of partners in agilesupply chain with help of information analysistools like Quality Function Deployment (QFD).

Most of the available systems are developed fortraditional manufacturing environments wherea static market behavior and resources have beenemployed for producing goods and services. Forphysically distributed manufacturing environ-ments, systems such as ERP would be more suit-able. The support systems for AM heavily rely oncomputer-based information systems includingEDI, Internet and Electronic Commerce. There-fore, a #exible architecture for systems to accom-modate temporary alliances will help to improveenterprise integration and hence agility in organ-izations.

2.2.4. PeopleThis section reviews the literature on human

factors in AM environments. The demands thatAM initiatives will place on the current and emerg-ing work force to achieve increasing levels of qual-ity and #exibility with lower costs and shorterproduct life cycles are discussed in this section. Theissues of worker selection, continuous skill develop-ment, work place design, equipment maintenance,process improvement, mistake proo"ng, and pro-cess recon"guration for new products are to beconsidered from the perspective of workforcerequired for the development of AMSs [57]. Thefollowing key issues of human factors are to be con-sidered in agile environment: knowledge workers,multilingual workforce, multinational workforce,incentive schemes, type and level of education andtraining, relation with unions, and pay award.

2.2.4.1. Knowledge workers. Agile manufacturingcan be operated e!ectively with the help of know-ledge workers such as computer operators, drafts-man, design engineers and maintenance engineer.Since AM is of more IT-intensive, there is a need toimprove the productivity of knowledge workerswith the objective to achieve agility in manufactur-ing. An appreciation of the human factors inherentin agile product development is pivotal to the


successful integration of agility-enabling technolo-gies, as well as the coordination of personnel work-ing within a CE environment. Forsythe [56] brie#ysummarizes human factors contributions to: (1) de-velopment of agile business practices; (2) design ofenabling technologies; and (3) management of theintroduction and "elding of new technologies andbusiness practices. The author also discussed hu-man factors related to the communications andinformation infrastructure essential to an organiza-tion making the transition from traditional to agileproduct development. Also, knowledge workersshould be #exible taking into account the respons-ibility of multiple tasks and communicating inmulti-lingual in a physically distributed environ-ment of AM enterprise.

2.2.4.2. Top management support and employee em-powerment. Gunasekaran [17] highlights the roleof employee empowerment in improving theco-operative supported work in a physically dis-tributed VE enterprises. Achieving agility in manu-facturing requires radical changes in the line ofreengineering business process. This level of changein any organization demands a total support of topmanagement in terms of providing necessary tech-nical and "nancial support together with employeeempowerment. As part of a multidisciplinary em-powered and self-directed teams, a human factorspractitioner can be asked to assume leadership ofthe communications team with the task of develop-ing an information infrastructure to support theproduct development project structure and thefacilitation of information #ow in a large, geo-graphically dispersed, project team. This responsi-bility should re#ect a recognition of the importanceof human interactions in an information-drivenproduct development process [59]. Also, top man-agement involvement is vital to e!ective re-engin-eering of the supply chain and logistics in agileenvironments [21]. There is not much discussion inthe literature concerning the top management sup-port and level of involvement in the design andimplementation of AMSs.

2.2.4.3. Training and education. It is not clear yetwhat sort of training and education required tomotivate the employees to take part in the develop-

ment of AM. This section spells out some of majortraining and education requirements in AM envi-ronments.

Agile manufacturing has di!erent requirementsof workforce as compared to that of traditionalsystems and they are: (i) closer interdependenceamong activities, (ii) di!erent skill requirements,usually higher average skill levels, (iii) more im-mediate and costly consequences of any malfunc-tion, (iv) output more sensitive to variations inhuman skill, knowledge and attitudes and to men-tal e!ort rather than physical e!ort, (v) continualchange and development, and (vi) higher capitalinvestment per employee, and favor employees re-sponsible for a particular product, part or process[60]. These to some extent de"ne the character-istics of agile workforce and the training and educa-tion required and some of them are: IT-skilledworkers, knowledge in team working and negoti-ation and advanced manufacturing strategies andtechnologies, empowered employees, multifunc-tional workforce, multi-lingual workforce, andself-directed teams [9,44,56,59]. In a VE, the natureof training and education should have a di!erentfocus as compared to that of traditional organiza-tions. For example, an international team of em-powered employees and self-directed teams shouldbe developed with a view to improve the e!ec-tiveness of a globally distributed manufacturingenterprise. This requires the understanding of theculture and language of each other together withsu$cient literacy in computerised informationanalysis and synthesis.

The literature available on AM workforce israther limited [17]. The reason for this is that thereis no clear cut framework for identifying the im-plications of AM on workforce characteristics, andmost of the literature deal with enabling technolo-gies and some strategies of AM. However, humanfactors play a signi"cant role in the successful de-velopment and implementation of AM.

3. Comments on agile manufacturing literature andfuture research directions

Theoretically derived hypotheses and empiricalstudies to test them are conspicuously absent from


studies of the agile organization. A study by Hoytet al. [62] a scene in this direction. Although itseems intuitive that the ability to respond to dy-namic and unpredictable changes in the environ-ment should contribute to a company's success, thisfact has not been scienti"cally tested.

The survey of literature has provided furtherinsights into the AM concepts and systems includ-ing strategies, technologies, systems and people.From the analysis of AM literature, the followingare some of the future research directions thatwould assist in achieving agility in manufacturingmore e!ectively [17].

(i) A methodology for evaluating potential part-ners for agile enterprises based on core-com-petencies and market forces needs to bedeveloped. Criteria for selecting partners ofagile enterprise should be identi"ed with thehelp of suitable conceptual and empirical re-search.

(ii) A framework for determining the type andlevel of di!erent skills required for agile enter-prise should be developed using multiple sitecase studies. Furthermore, the nature of train-ing and education required for workforceshould be precisely de"ned taking into ac-count the geographically dispersed partnersin AM environments. Suitable informationsystems need to be developed for determiningthe type of skills and numbers of workersrequired in agile environments.

(iii) The infrastructure and organizational charac-teristics of agile enterprise can be determinedby developing a suitable theoretical frame-work and testing them with real life manu-facturing environments. This frameworkessentially centers around the nature of in-formation and material #ows in AM enter-prise. The issues of temporary alliances basedon core competencies are to be embedded inthe proposed empirical studies.

(iv) In agile enterprises, supply chain linksare often temporary and hence #exible.Therefore, there is need to develop suit-able performance measurements and in-vestment justi"cation techniques for thisenvironment.

(v) An investigation on the selection of suitablearchitectures for agile enterprises would o!erfurther insights into the design of AMSs. Also,appropriate capacity planning and schedul-ing methods are to be devised to support thee!ective operations of physically distributedvirtual enterprises in AM environments.

(vi) In a physically distributed agile enterpriseenvironment, there is a need for a di!erentquality management system. All the modernquality management strategies and methodscan be used for agile environments, but needto be modi"ed taking into account the recon-"gurability and dynamics of agile organiza-tions.

(vii) The cost accounting systems such as Activity-Based Costing (ABC) would be suitablefor advanced manufacturing environments.However, considering the characteristics ofagile enterprise, the application of ABC needsfurther investigation. Physically distributedmanufacturing environment demands a simplecost accounting system to overcome the di$-culties of communication, integration anddomestic regulations among geographicallydispersed partners.

(viii) Gaining rapid response to changing customerdemand requires equally agile logistics.Logistics can be helped by appropriateproduct design and tooling to ease mater-ials handling. Fewer stock lines and greaterinterchangeability of items reduce the work-ing capital and the risk of obsolescence ofslow moving lines [21]. The issue of logis-tics in AM environments has not receivedsigni"cant attention from researchers. Forexample, what sort of systems would besuitable for purchasing and distribution ofgoods would be appropriate in AM envi-ronments which include VE based ontemporary alliances need to be investigatedfurther.

Realizing the importance of a framework fordeveloping AMSs, an attempt has been made inthis paper to develop a framework for AM basedon the literature survey and other reported caseexperiences.


4. A framework for the design of agilemanufacturing systems

In this section, a framework for the design ofAMSs is developed. This development is based onthe literature survey and its analysis. It can be seenthat most of the literature on AM and related issueseither deal with strategies or techniques, but not anintegrated view of developing an AMS. In thissection, an attempt has been made to present anintegrated strategic and techniques framework forthe design and development of AMSs together withpeople and systems issues. Agile manufacturing cansuccessfully be accomplished using various well-de-"ned agile system architectures. The system archi-tecture for AM may include control, function,process, information, communication, distribution,development, and implementation [46].

E!ective and e$cient implementation of AMSsrequire enterprise level integration. The "rst step inthis direction is to integrate design, process plann-ing and scheduling. A bidding-based approach tothe integration of computer-aided design, processplanning and real time scheduling can be used forthe design and implementation of AMS [63]. Theproduct is represented in a STEP model with de-tailed design and administrative information in-cluding design speci"cations, batch size, and duedates. Upon arrival at the manufacturing facility,the product is registered in the shop #oor managerwhich is essentially a coordinating agent. The shop#oor manager broadcasts the product's require-ments to the machines. The shop contains auto-nomous machines that have knowledge about theirfunctionality, capabilities, tooling and schedules.Each machine has its own process planner andresponds to the product's request in a way that isconsistent with its capabilities and capacities.When more than one machine o!ers certainprocess(es) for the same requirements, they enterinto negotiation. Based on processing time, duedate and cost, one of the machines wins the con-tract. The successful machine updates its scheduleand advises the product to request raw material forprocessing. The concept was implemented usinga multi-agent system in an object-oriented pro-gramming language. The task of decompositionand planning are achieved through contract nets

[63]. As discussed earlier, Internet plays a signi"-cant role in AMSs.

Based on the literature survey, a conceptualmodel for the development of AMSs is developed asshown in Fig. 1. As indicated earlier, the model hasbeen developed along four key dimensions includ-ing strategies, technology, people and systems. Themain objective here is to develop an integratedAMSs with the help of suitable strategies and tech-niques to develop rapid partnership formation, VEand recon"gurability for mass customization. Fur-ther details of the model are discussed hereunder.

4.1. Strategies

Long-term decisions considering the recon-"gurability of the organization with the objective tocompete in the global market by mass customiz-ation are important to make use of various re-sources available for producing quality goods andservices. The following are some of the strategicdecision areas that should be considered while de-veloping AMSs: CE, rapid partnership formation,strategic alliances, virtual enterprise, physically dis-tributed manufacturing systems [17].

Partner selection is one of the most importantactivities in agile enterprise. Selecting manufactur-ing partners in AM is an endeavour that is worriedwith the complexity and dynamic of the marketdriven by customer needs as well as the inherentsubjectivity of the selection process. Traditionalvendor selection methodologies do not lend them-selves as a ready solution to these needs of agileenvironment [11]. The criteria for selecting part-ners are not only based on the cost and responsive-ness, but should also be based on the quality ofgoods and services, location of the company, ITskills, and the e!ectiveness of their supply chainincluding the #exibility. For instance, CE can beused for new product developments and partnerselection based on core competencies.

The increased customization, and the subsequentadoption of CE practices driven by the signi"canceof the product time-to-market and compounded bythe distributed nature of enterprises, makes it verydi$cult to accurately estimate the product manu-facturing cost and the cycle time for new products.


Fig. 1. Development of an agile manufacturing system.

Appropriate cost accounting practices and perfor-mance measures should be adopted in AM environ-ments. It is not simply based on non-"nancialperformance measures, but also "nancial perfor-mance measures. For example, the partner's perfor-mance can be measures by the time to deliver theproduct and of course the cost as well. The intan-gible factors are signi"cant in AM environmentsand are di$cult to quantify, such as the in-houseinformation and communication systems. There-fore, there is a need to identify and measure intan-gibles in agile enterprises [20].

In AM environments, prequalifying partners,evaluating a product design with respect to thecapabilities of potential partners, and selecting theoptimal set of partners for the manufacture of a cer-tain product are important. A decision supportsystem for design evaluation and partner selectionin AM needs to be developed. Therefore, everycompany must "nd the right combination of cul-ture, business practices, and technology that arenecessary to achieve agility in its organization.After analyzing the strategic options of AM, the

strategies such as market forces, partnership forma-tion, VE, rapid product developments are the keystrategies to achieve agility in manufacturing.

4.2. Technology

As discussed earlier, strategies of AM with suit-ably supported technologies would enhance thesuccess of achieving/improving agility in organiza-tions. Therefore, key technologies to achieve agilitybased on the strategies selected are highlightedhere. In AMSs, rapid hardware changeover can bemade possible by robots, #exible part feeders, #ex-ible "xturing, modular grippers and modular as-sembly hardware. The #exible feeders rely on beltfeeding and binary computer vision for pose es-timation. This has a distinct advantage over non-#exible feeding schemes such as bowl feeders whichrequire considerable adjustment to changeoverfrom one part to another [55].

The agile-enabling technologies such as Internet,Multimedia, EDI, Electronic Commerce, #exible


manufacturing cells, Robotics and CAD/CAMneed to be suitably incorporated within the scopeof the VE to achieve agility in manufacturing.

4.3. Systems

In the AM paradigm, where multiple "rms co-operate under #exible virtual enterprise structures,there exists a great need for a mechanism to man-age and control information #ow among collab-orating partners. In response to this pressing need,an AM information system integrating manufac-turing databases dispersed at various partner sitesneeds to be developed.

Agile manufacturing requires agile support sys-tems which can be obtained by computer-aidedinformation systems for planning and control activ-ities of manufacturing. Systems such as MRP, ERP,and CAD/CAM and KBS can be used to collectinformation and make appropriate decisions con-cerning the e!ective operations that would supportagility in manufacturing organizations. The reasonwhy computerized information systems are recom-mended for AM is due to the characteristics such asVE, rapid prototyping and physically distributedmanufacturing systems.

There is a need for the development and integra-tion of information infrastructure to facilitate dis-tributed design, planning, manufacturing, andmarketing functions in agile enterprise. The pur-pose of this is to achieve multipath agility in theproduct development process. Multipath agilityprovides increased access to alternative resourcesand information, and achieves improvements inproductivity and quality through #exibility of accessand utilization of resources, rather than throughstepwise improvements in any one task. These con-cepts will be tested through demonstration casestudies in many-to-one and one-to-many design}manufacturing cycles of real products [42]. Never-theless, integration of current fragmented computersystems, causing over-complexity, is perhaps thebiggest challenge the AM enterprise faces [36].

Partner organizations may remain competitivein some markets or return to a competitive rela-tionship with other partners after the dissolution ofthe VE. Sensitive information and business pro-

cesses must be protected as part of the overallbusiness process design, so as to enable the type ofcommunication necessary to allow the partners toperform successfully as a single entity. Computerizedinformation systems with suitable protection mayhelp to safeguard sensitive information [17].

To make a transition to the agile industry, wheremanufacturing "rms cooperate under virtual enter-prise structures, there is a need to develop mecha-nisms for agile partners to cooperate #exibly anddynamically. An individual partner, with its in-formation system, can plug in (out) of the VE sup-ported by a communication network such as theInternet. It is through the integration of these part-ner information systems that information manage-ment is possible in collaborative activities. Theconceptual information system for an agile manu-facturer consists of two hierarchies, viz. the in-formation hierarchy and the transaction hierarchy.The information hierarchy represents agile in-formation using Object Oriented Methodology(OOM), and the transaction hierarchy models part-ner transaction management using a KnowledgeBased Systems (KBS) accommodating expert rules.Using the Client}Server architecture, an integratedpartner information system for improving #exiblyand dynamically can be developed [15].

In agile enterprise where manufacturing partnersshare product related data to come up with new,customized, and high quality products at minimalleadtimes. The principles of group technology (GT)can be used. The information on product design isassumed to be available in the product databases ofdistributed partners, and can be generated throughan existing GT design processor. A software systemusing object-oriented technology will be useful inimplementing the procedure [48]. Agile manufac-turing systems can bene"t signi"cantly from adatabase support. An AM database system(AMDS) designed for capturing and manipulatingthe operational data of a manufacturing cell will behelpful.

Design for agile assembly is accomplished byconsidering operational issues of assembly systemsat the early product design stage. The #exibilityrequired of an AMS must be achieved largelythrough computer software. The system's controlsoftware must be adaptable to new products and to


new system components without becoming unreli-able or di$cult to maintain. This requires design-ing the software speci"cally to facilitate futurechanges [25].

Agile manufacturing can be successfully accomp-lished using various well-de"ned system architec-tures. For example, Jung et al. [46] provide aprimary sketch of architectural requirements forrapid development of AMSs. This architecture caninclude control, function, process, information,communication, distribution, development, andimplementation.

4.4. People

The most critical problem in agile environment ishow to manage and motivate workforce to supportthe #exibility and responsiveness. Forsythe [56]discusses the human factors in AM. For the devel-opment of agile business practices, there needs to beconsideration of human factors a!ecting decisionmaking within fast-paced dynamic environments[58]. When information does not #ow, due to tech-nical or human issues, agility is lost. For thisreason, elimination of human points of failure ininfrastructure support is essential [59]. If users areunwilling or reluctant to accept agile practices andenabling technologies, AM will fail from the inabil-ity to overcome the inertia of traditional, oftendeeply ingrained practices. Agile manufacturingposes threat to the comfort of managers due to theempowerment of product development teams andthe increased openness of information.

Concurrent engineering within a fast-pacedproduct development environment favors collab-orative work between engineering disciplines. Mostof the challenges of human factors posed by theagile environment can be overcome by a series ofteam meetings during which the team jointlydeveloped the project plan, including objectives,strategies for meeting objectives, a detailed tasknetwork, schedule and resource and funding pro-jections. The information technologies alone arenot su$cient to achieve the desired communica-tions e$ciency and, if anything, the unfamiliarity ofthe technologies could impede communications ef-"ciency. To overcome these factors, two strategies

can be adopted: (i) various mechanisms can beemployed to allow team members to meet face-to-face and establish personal familiarity (e.g. teamtraining and project planning), and (ii) to sustainfamiliarity throughout the course of the project(through monthly all-hand meetings, weeklylunches, social events).

Information may be transmitted via multiplechannels depending on urgency, content and distri-bution through phone, voice-mail, fax, e-mail, andhttp [56]. The agile workforce should be capable ofmeeting increasing technological challenges, de-signing their work places, solving quality-relatedproblems and team-to-team learning, improvingequipment availability, using mistake-proo"ngprocesses, dealing with increased complexity, and"nally, helping labor unions harmonise their mem-bers and company expectations [57].

5. Summary and conclusions

With the rapid changes taking place in the globalmarket, it becomes clear that enterprises workingon an AM base or in mass customization willrapidly become leaders. In order to design e$cientproduction systems and factories operating in thesemanufacturing paradigms, new design approachesneed to be developed. This paper synthesizes someof the research work done on the AM to developa framework for the development of AMSs. Themodel presented in this paper can be tested with thehelp of suitable empirical and multiple case studies.Also, classical organization theory and strategy re-search methods are useful for the characterizationof AM organizations. The subject of AM needsmore feasibility studies from the perspectives ofestablishing VE, temporary alliances and their im-plications on the relationship with labor unionsand jurisdictions. The following are some of themajor issues that could be considered by re-searchers and practitioners when developingAMSs:

(i) The relationship between competitive basissuch as #exibility, quality, productivity,responsiveness, and cost needs to be investi-gated in relation to agility in manufacturing.


(ii) Knowledge management is a crucial issue ina VE enterprise and therefore, a frameworkshould be developed for the e!ective manage-ment of knowledge in AM enterprises.

(iii) There are not many studies on the applicationof operations research models for the designand development of AM. However, suitablemodels need to be developed for the purpose ofmaking decisions in a virtual and physicallydistributed manufacturing environment.

(iv) The workforce characteristics in AM need tobe de"ned more precisely with the help of ap-propriate empirical and case study research.

The following are the summary of issues thatshould be addressed in agile manufacturing as com-pared to traditional advanced manufacturing sys-tems: (i) the implications of temporary alliances onthe enterprise communication and co-ordination,(ii) the in#uence of virtual enterprise and physicallydistributed manufacturing on the human relationsmanagement, and (iii) the technologies and humanskills required for the information intensive manu-facturing environment.

Acknowledgements

The author is most grateful to two anonymousreferees for their helpful comments on the earlierversion of the manuscript which helped to improvethe presentation of the paper considerably.

References

[1] P. Bunce, P. Gould, From Lean to Agile Manufacturing,IEE Colloquium (Digest) (278) (1996) 3/1}3/5.

[2] P. Gould, What is agility, Manufacturing Engineer 76 (1)(1997) 28}31.

[3] S.M.R. JamesMoore, Agility is easy, but e!ective agilemanufacturing is not, IEE Colloquium (Digest) (179)(1996) 4.

[4] H. Cho, M. Jung, M. Kim, Enabling technologies of agilemanufacturing and its related activities in Korea, Com-puters and Industrial Engineering 30 (3) (1996) 323}334.

[5] P.T. Kidd, Agile manufacturing: a strategy for the 21stcentury, IEE Colloquium (Digest) (74) (1996) 6.

[6] R. DeVor, R. Graves, J.J. Mills, Agile manufacturing re-search: Accomplishments and opportunities, IIE Transac-tions 29 (10) (1997) 813}823.

[7] R. Booth, Agile manufacturing, Engineering ManagementJournal 6 (2) (1996) 105}112.

[8] M.F. McGrath, Making supply chains agile for nicheproducts, Annual International Conference Proceedings} American Production and Inventory Control Society,1996, pp. 167}171.

[9] R.A. Abair, Agile manufacturing: This is not just repackag-ing of material requirements planning and just-in-time,Annual International Conference Proceedings } AmericanProduction and Inventory Control Society, 1995,pp. 196}198.

[10] S.K. Gupta, J.W. Herrmann, G. Lam, I. Minis, Automatedhigh level process planning for agile manufacturing,American Society for Mechanical Engineers, DesignEngineering Division (Publication) DE 83 (2) (Pt 2) (1995)pp. 835}852.

[11] P. Gupta, R. Nagi, Flexible optimization frameworkfor partner selection in agile manufacturing, IndustrialEngineering Research } Conference Proceedings, 1995,pp. 691}700.

[12] J.W. Herrmann, I. Minis, V. Ramachandran, Informationmodels for partner selection in agile manufacturing, Com-puter Engineering Division, CED, American Society ofMechanical Engineers 7 (1995) 75}91.

[13] S. Jain, Virtual factory framework: A key enablerfor agile manufacturing, IEEE Symposium on EmergingTechnologies & Factory Automation 1 (1995)247}258.

[14] J.J. Mills, Integrated information infrastructure foragile manufacturing, American Society of Mechanical En-gineers, in: Manufacturing Engineering Division, MED,vols. 2}2, ASME, New York, NY, USA, 1995, pp.1037}1047.

[15] L. Song, R. Nagi, Design and implementation of a virtualinformation system for agile manufacturing, IIE Transac-tions 29 (10) (1997) 839}857.

[16] Y. Tu, Production planning and control in a virtualone-of-a-kind production company, Computers in Industry34 (3) (1997) 271}283.

[17] A. Gunasekaran, Agile manufacturing: Enablers and animplementation framework, International Journal of Pro-duction Research 36 (5) (1998) 1223}1247.

[18] M.J. Tracy, J.N. Murphy, R.W. Denner, B.W. Pince, F.R.Joseph, A.R. Pilz, M.B. Thompson, Achieving agile manu-facturing: Part II, Automotive Engineering 102 (12) (1994)13}17.

[19] E.D. Adamides, Responsibility-based manufacturing, In-ternational Journal of Advanced Manufacturing Techno-logy 11 (6) (1996) 439}448.

[20] R.J. Graves, A. Agrawal, K. Haberle, Estimating tools tosupport multipath agility in electronics manufacturing,IEEE Transactions on Components, Packaging, andManufacturing Technology Part C: Manufacturing 19 (1)(1996) 48}56.

[21] M. Pellew, Agile manufacturing } the role of rapidresponse logistics, IEE Colloquium (Digest) (179)(1996) 3.


[22] L.L. Hessney, Global supply chain management for thenext millennium: Fundamentals#integration"break-through, Annual International Conference Proceedings} American Production and Inventory Control Society,1997, pp. 387}389.

[23] J.H. Graham, R.K. Ragade, Design support system foragile manufacturing, Proceedings of the IEEE Interna-tional Conference on Systems, Man and Cybernetics1 (1994) 512}517.

[24] B.P. Erdel, Advanced machining processes integrate agilemanufacturing, SAE Special Publications 1233 (1997)115}122.

[25] A. Kusiak, D.W. He, Design for agile assembly: An opera-tional perspective, International Journal of ProductionResearch 35 (1) (1997) 157}178.

[26] S.S. Medhat, J.L. Rook, Concurrent engineering } pro-cesses and techniques for the Agile Manufacturing Enter-prise, IIE Conference Publication, (435) (1997) 9}14.

[27] M. Hong, S. Payandeh, W.A. Gruver, Modelling andanalysis of #exible "xturing systems for agile manufac-turing, Proceedings of the IEEE International Confer-ence on Systems, Man and Cybernetics 2 (1996)1231}1236.

[28] J.E. Spencer Jr., Robotics technology and the advent ofagile manufacturing systems in the footwear industry,Assembly Automation 16 (3) (1996) 10}15.

[29] S. Ashley, Rapid-response, Mechanical Engineering 119(12) (1997) 72}74.

[30] D. Enke, C. Dagli, Automated misplaced component in-spection for printed circuit boards, Computers and Indus-trial Engineering 33 (1}2) (1997) 373}376.

[31] G.H. Lee, Recon"gurability consideration for the design ofcomponents and manufacturing systems, InternationalJournal of Advanced Manufacturing Technology 13 (5)(1997) 376}386.

[32] R.G. Mathieu, Manufacturing and the Internet, AnnualInternational Conference Proceedings } American Pro-duction and Inventory Control Society, APICS, FallsChurch, VA, 1997, pp. 368}370.

[33] F.L. Merat, N.A. Barendt, R.D. Quinn, G.C. Causey, W.S.Newman, V.B. Velasco Jr., A. Podgurski, Y. Kim, G.Ozsoyoglu, J.Y. Jo, Advances in agile manufacturing, Pro-ceedings } IEEE International Conference on Roboticsand Automation, vol. 2, 1997, IEEE, Piscataway, NJ, USA,pp. 1216}1222.

[34] E.A. Orady, T.A. Osman, C.P. Bailo, Capability study ofrobotics and manufacturing cell simulation software,Computers & Industrial Engineering 33 (1}2) (1997)83}86.

[35] R.D. Quinn, G.C. Causey, F.L. Merat, D.M. Sargent, N.A.Barendt, W.S. Newman, V.B. Velasco Jr., A. Podgurski,J.Y. Jo, L.S. Sterling, Y. Kim, Agile manufacturing work-cell design, IIE Transactions 29 (10) (1997) 901}909.

[36] P. Bocks, Enterprise data management framework foragile manufacturing, Computer Engineering Division,CED, American Society of Mechanical Engineers, NewYork, vol. 7, 1995, pp. 41}46

[37] Z.Y. Wang, K.P. Rajurkar, A. Kapoor, Architecture foragile manufacturing and its interface with computer integ-rated manufacturing, Journal of Materials ProcessingTechnology 61 (1}2) (1996) 99}103.

[38] X. Yang, Object-oriented model of agile manufacturingsystem, Proceedings of the IEEE International Conferenceon Industrial Technology, IEEE, Piscataway, NJ, USA,1996, pp. 171}175.

[39] J.Y. Jo, Y. Kim, A. Podgurski, W.S. Newman, Virtualtesting of agile manufacturing software using 3D graphicalsimulation, Proceedings } IEEE International Conferenceon Robotics and Automation, 2 (1997) 1223}1228.

[40] S. Kirk, E. Tebaldi, Design of robotic facilities for agileautomobile manufacturing, Industrial Robot 24 (1) (1997)72}77.

[41] A. Kusiak, C.X. Feng, Design of products for an agilemanufacturing environment, Computers in Industrial En-gineering, Proceedings of the International Conferenceand Exhibit, 1, 1994, pp. 413}420.

[42] A.C. Sanderson, R.J. Graves, D.L. Millard, Multipath agil-ity in electronics manufacturing, Proceedings of the IEEEInternational Conference on Systems, Man and Cybernet-ics, vol. 1, 1994, pp. 501}505.

[43] A. Candadai, J.W. Herrmann, I. Minis, Group tech-nology-based variant approach for agile manufac-turing, American Society of Mechanical Engineers,Manufacturing Engineering Division, MED 1 (1995)289}305.

[44] C. Forsythe, M.R. Ashby, User-driven product data man-ager system design, Proceedings of the Human Factorsand Ergonomics Society 2 (1995) 1170}1174.

[45] S. Iyer, R. Nagi, Identi"cation and ranking of similar partsin agile manufacturing, Industrial Engineering Research} Conference Proceedings (1995) pp. 709}717.

[46] M. Jung, M.K. Chung, H. Cho, Architectural requirementsfor rapid development of agile manufacturing systems,Computers & Industrial Engineering 31 (3}4) (1996)551}554.

[47] D.W. He, A. Kusiak, Design of assembly systems formodular products, IEEE Transactions on Robotics andAutomation 13 (5) (1997) 646}655.

[48] S. Iyer, R. Nagi, Automated retrieval and ranking of sim-ilar parts in agile manufacturing, IIE Transactions 29 (10)(1997) 859}876.

[49] Y. Kim, J.Y. Jo, V.B. Velasco Jr., N.A. Barendt, A.Podgurski, G. Ozsoyoglu, F.L. Merat, Flexible softwarearchitecture for agile manufacturing, Proceedings } IEEEInternational Conference on Robotics and Automation4 (1997) 3043}3047.

[50] L. Monplaisir, AI enhancements for a CSCW for manufac-turing systems design, Industrial Engineering Research} Conference Proceedings, 1997, IIE, Norcross, GA, USA,1997, pp. 158}163.

[51] S. Lee, H.C. Kuo, N.H. Balkir, G. Ozsoyoglu, Databaseserver architecture for agile manufacturing, Proceedings} IEEE International Conference on Robotics and Auto-mation, vol. 4, IEEE, Piscataway, NJ, 1997, pp. 3048}3053.


[52] A. Candadai, S. Champati, J.W. Herrmann, I. Minis, V.Ramachandran, Information needs in agile manufacturing,ASME Database Symposium, 1994, pp. 101}109.

[53] U.G. Gupta, R.O. Mittal, Quality, time, and innovationbased performance measurement system for agile manu-facturing, Proceedings } Annual Meeting of the DecisionSciences Institute, vol. 3, 1996, pp. 1511}1513.

[54] T.B. McMullen Jr., Theory of constraints: The infrastruc-ture of agile manufacturing, Annual International Confer-ence Proceedings } American Production and InventoryControl Society, 1996, pp. 336}346.

[55] R.D. Quinn, G.C. Causey, F.L. Merat, D.M. Sargent, N.A.Barendt, W.S. Newman, V.B. Velasco Jr., A. Podgurski,J.Y. Jo, L.S. Sterling, Y. Kim, Design of an agile manufac-turing workcell for light mechanical applications, Proceed-ings } IEEE International Conference on Robotics andAutomation, vol. 1, 1996, pp. 858}863.

[56] S. Forsythe, Human factors in agile manufacturing: A briefoverview with emphasis on communications and informa-tion infrastructure, Human Factors and Ergonomics inManufacturing 7 (1) (1997) 3}10.

[57] F.E. Plonka, Developing a lean and agile work force,International Journal of Human Factors in Manufactur-ing 7 (1) (1997) 11}20.

[58] K.M. Eisenhardt, Making fast strategic decisions in high-velocity environments, Academy of Management Journal32 (3) (1989) 543}576.

[59] C. Forsythe, M.R. Ashby, Human factors in agile manufac-turing, Ergonomics in Design 4 (1) (1996) 15}21.

[60] A. Pinochet, Y. Matsubara, M. Nagamachi, Constructionof a knowledge-based system for diagnosing the sociotech-nical integration in advanced manufacturing technologies,The International Journal of Human Factors in Manufac-turing 6 (4) (1996) 323}349.

[61] R.A. Abair, Agile manufacturing: successful implementa-tion strategies, Annual International Conference Proceed-ings } American Production and Inventory ControlSociety, 1997, pp. 218}219.

[62] J. Hoyt, F. Hug, D.H. Liles, Design of a study to test thestatistical signi"cance of "ve enablers of agile manufactur-ing organizations, Industrial Engineering Research }Conference Proceedings (1997) 966}971.

[63] P. Gu, S. Balasubramanian, D.H. Norrie, Bidding-basedprocess planning and scheduling in a multi-agent system,Computers and Industrial Engineering 32 (2) (1997)477}496.


Documents

Tele Manu Fatur a 72