Design of effective ework

8/14/2019 Design of effective ework

1/24

Production Planning & Control ,Vol. 14, No. 8, December 2003, 681703

Design of effective e-Work: review of models, tools,and emerging challenges

SHIMON Y. NOF

Keywords active protocols, active middleware, autonomousagents, collaborative robotics, collaborative work, conict reso-lution, distributed operations, enterprise integration, enterprisemodels, parallelism, teamwork evaluation, viability

Abstract. The foundations and scope of e-Work are described,and investigations of fundamental design principles for e-Workeffectiveness reviewed. The premise is that without effectivee-Work, the potential of emerging and promising electronicwork activities, such as virtual manufacturing, teleroboticmedicine, automated construction, intelligent transportation,and e-business, cannot be fully materialized. A typical andrecent example of the inability to fulll the potential of e-Work is the frustration of workers over supply chains/networks with complex ERP and other information systems,originally designed to simplify and improve their performance.Challenges and emerging e-Work solutions are described,and recent discoveries are clustered in four areas, e-Work;Integration, Coordination and Collaboration; DistributedDecision Support; and Active Middleware, with annotatedreferences. PRISM Center developments of e-Work designprinciples, models and tools are described, including: Coopera-tion Requirement Planning; the principle of parallelism; theprinciple of conict resolution; new measures of viability andscalability; and the Teamwork Integration Evaluator (TIE),which applies the analogy of distributed computing.

1. Introduction: denitions and scope of e-Work

As power elds, such as magnetic elds and gravita-tion, inuence bodies to organize and stabilize, so does

the sphere of computing and information technologies. Itenvelops us and inuences us to organize work systems ina different way, and purposefully, to stabilize work whileeffectively producing the desired outcomes.

e-Work was dened by the PRISM Center (Nof 1999,

2000a, 2000b) as any collaborative, computer-supportedand communication-enabled productive activities inhighly distributed organizations of humans and/or robotsor autonomous systems (gure 1). In essence, e-Workis comprised of e-activities, namely, activities based onand executed through information technologies. Ingure 1, those e-activities include v-(virtual)Design,e-Business, e-Commerce, Intelligent Robotics,e-Manufacturing, v-(virtual)Factories, e-Logistics,i-(intelligent) Transportation, and v-(virtual)Enterprises.All these e-activities rely on computer support and com-munication technologies, and all of them require collab-oration in their inherent interactions between machine,humans, and computers. Other examples of e-activitiesthat comprise e-Work are telemedicine, e-Learning,e-Training, navigation and security monitoring by geo-graphic information systems (GIS), and more. e-Workincludes applications such as telerobotics (e.g. Brady andTarn 2000, Hamel 2000) and remote medical and healthservices (e.g. Meng et al. 2000, Rovette 2000). Some havedened e-Work as Telework (Chiozza and Stanford-Smith 2001, Burgess 2003) and indeed, Telework can bepart of e-Work according to the general denition above.

Author : Shimon Y. Nof, PRISM Center Production, Robotics, and Integration Software forManufacturing & Management, Purdue University, West Lafayette, Indiana, 47907-2023, USA,E-mail: [email protected]

SHIMON Y. N OF , Professor of Industrial Engineering at Purdue University, has held visiting posi-tions at MIT and Universities in Chile, EU, Hong Kong, Israel, Japan, and Mexico. Director of the NSF-industry supported PRISM Center for Production, Robotics, and Integration Softwarefor Manufacturing & Management; he is a Fellow of IIE, Secretary General of IFPR, and currentChair of IFAC CCManufacturing & Logistics Systems. He has published over 250 articles onproduction engineering and information/robotics engineering and management, and is the author/editor of nine books in these areas. In 1999 he was elected to the Purdue Book of Great Teachers,and in 2002 he was awarded the Engelberger Medal for Robotics Education. Nof has also had overeight years of experience in industrial positions.

Production Planning & Control ISSN 09537287 print/ISSN 13665871 online # 2003 Taylor & Francis Ltdhttp://www.tandf.co.uk/journals

DOI: 10.1080/09537280310001647832


2/24

Several conferences on e-Work as Telework have takenplace since 2000 (Cheshire Henbury 2003). Increasingly,however, it is recognized that human work has beenand still is being fundamentally transformed, as it hasbeen throughout history, by the combination of socialorganization and technology advancements (table 1).

The principles of the work system have been trans-formed, and the focus of concern in the design of worksystems has changed. While advancement in technologyover time has increased the ability to work and produce,the challenges and complexities have also increased, and

seem to be increasing at an exponential rate. As shownin table 1, the present challenge with e-Work systems isto reduce, and possibly optimize both the informationoverload and the task overload.

The objective of this article is to describe the founda-tions and scope of e-Work, and review investigations of fundamental design principles for e-Work effectiveness.The premise is that without effective e-Work, the poten-tial of emerging and promising electronic work activities,such as virtual manufacturing, telerobotic medicine,automated construction, intelligent transportation, ande-business, cannot fully materialize. A typical and recentexample of the inability to full the potential of e-Work isthe frustration of workers over supply chains/networkswith complex ERP and other information systems, orig-inally designed to simplify and improve their perfor-mance. Even if the humancomputer interface for suchsystems is highly effective, the information overload and

expectations from the human operators can be over-whelming. Effective production of the best outcomesmay not occur.

The view of e-Work as the foundation of e-activities(gure 1) implies that there are work abilities that can besignicantly better enabled by information technologies,for example:

. Better connectivity and ability for farther globalreach.

. Enhanced communication and coordination.

. Acceleration of knowledge sharing and distribution.

. Better interactivity, exibility, customization-ability.

Definition [PRISM Center, 1999]

Collaborative, computer-supported, andcommunication-enabled operations in highlydistributed organizations of humans / robots / autonomous systems

Our goal: Augment human abilities to work

Challenges : Complexity Dependence Integrity Communication Coordination Noise Mismatch

ee -- W o rW o r k

v-Design e-Businesse- CommerceI -

Roboticsv-

Enterprise ... ..e-Mfg

v-Factory

e-Logisticsi-Tran

e-Work

P R I S M La b / P u r d u e

Figure 1. The original denition of e-Work, its goal andchallenges (Nof 1999, 2000a, 2000b) (v-Design; Virtual

Design; I-Robotics: Intelligent Robotics; v-Factory:Virtual factory; i-Tran: Intelligent Tranportation;

v-Enterprise: Virtual Enterprise).

Table 1. Fundamental changes in the evolution of work systems.

PeriodTechnology

advancementHuman workaugmented by

Fundamentalwork system

principle

Engineering &management

concerns Work system goal 1

Before theIndustrialRevolution

Hand-tools Tooling Manual andanimal power

Work methods Enable workfunctions

BeforeComputers

Engines;Machines

Power;Motions; Moves

Human-machinesystems

Work ow Reduce muscleload

ComputerAge

Computers;Communication;Robots

Control;Data processing;Automation

Computer-aidedand computer-integrated systems

Humancomputerinteraction;Humanrobotinteraction

Reduce informationoverload

InformationAge

Telecom;Internet;Mobility

Cognitive skills;Collaboration

e-Work e-Work design;Collaboration;Parallelism

Reduce informationand task 2 overload

1 Goals beyond productivity, quality, safety, etc.2Both cognitive and non-cognitive tasks

682 S. Y. Nof


3/24

. Higher velocity of work tasks and exchanges.

. Reduced communication costs.

. Lower cost of work transactions.

. Lower agency cost by replacing humans with soft-ware agents (eliminating the middle people).

Hence, e-activities enhance the opportunities for better,more productive work, and at the same time, they widenthe competition. But despite the advantages entailed withe-Work, there are common challenges:

. Greater work complexity.

. More limitations caused by increasing interdepen-dence.

. Issues of integrity and trust.

. Greater need for coordination, cooperation, andsynchronization.

. Communication challenges and failures, complicat-ing systems requirements.

. Problems of mismatch, e.g. inconsistent versions,cultural differences, etc., also complicating systemsrequirements.

. New users training requirements and associated

costs.. The infancy stage of research, leading to applicabil-

ity limitations, and to hesitation in acceptance bypotential users.

Failing to overcome these common challenges mayexplain why often, e-activities do not full the expecta-tion promised by technology.

How is e-Work different from e-Business ande-Commerce? The same way work is different from busi-ness and commerce. They are highly related, but notthe same. Figure 2 depicts some of the scope differences.

e-WORK

e-WORK FEATURES &

FUNCTIONS

e-Operations Human-Computer

Interaction Human-Robot Interaction Integration Collaboration Coordination Networking e-Learning e-Training

e-WORK MODELS & TOOLS

Agents

Protocols Workflow Middleware Parallelism Teamwork Groupware

e-COMMERCE

CUSTOMERS(Individuals;

Other enterprises)

Online Orders Online Sales Financial

Transactions Design

customization Sales-force

Automation

SUPPLIERS

Online Purchasing Supply Chain

Management

Funds Transfer

e-BUSINESS / e-ENTERPRISE

FACTORIES, PLANTS

e-Design e-Mfg. ERP Robotics Facilities Process Automation

OFFICES, SERVICES

Team Projects Telework CRM Group Scheduling

ALLIANCE PARTNERS

Joint Projects Outsourcing Joint marketing e-Supply

OurEnterprise

OTHER e-ACTIVITIES

Examples: Telemedicine, e-Charities, e-Explorations, etc.

Figure 2. e-Work is fundamental to e-Business, e-Commerce, and other e-activities (just as work is fundamental to business,commerce, and other activities).

Design of effective e-Work 683


4/24

One way to dene the distinction between these threerelated areas is:

e-Commerce: Commercial e-exchange among customersand providers (suppliers). This exchange ismostly between separate, external entities.

e-Business: Same e-exchanges as in e-Commerce, exceptthe emphasis is on internal e-operationsand e-services within the scope of a givenenterprise.

e-Work: e-operations and e-support to enable theabove two, as well as many non-commercialand non-business work activities, from crea-tive artistic projects, education and learn-ing, to medical help, charity activities,police work, scientic exploration, andmany more productive work endeavors.

The main reason to distinguish e-Work is the realizationthat there are certain common principles in the designof effective e-Work, without which e-activities cannotsucceed. It can be said that To do good e-Business,somebody has to do good e-Work (Nof 2001).

e-Work depends on communication, includingtelecommunication. The Internet has inuenced manyof the tele-operation areas, yet there are e-Work activitiesthat clearly do not depend on the Internet. Table 2

shows examples of Internet-based e-Work and non-Internet based e-Work, as well as some that are enabledby both.

2. Major challenges and emerginge-Work solutions

During the industrial ages Before Computers(Table 1),HumanMachine Systems challenged engineers andmanagers to solve interference issues, namely, how tocoordinate and synchronize human workers, machineoperations, and material ow in and between factories.Planning and control systems were mostly manual.

With the Computer Age, new concerns emerged, increas-ingly, for better Interaction and for Integration (fromislands of automation to total systems, from local optimi-zation to global optimization). These concerns reectedthe challenges of new work abilities to interface morecomplex systems, involving more information transfers,many more transactions, and the dawn of computer androbot servitude: Beginning the delegation of controland cognitive tasks from humans to computers, machines,and robots. The area of HumanComputer Interaction(HCI) joined the areas of Industrial and Systems

Table 2. Examples of e-work cases with telecommunication.

e-Work enabled by communication Non-internet based Internet based Combination

Distributed computing and information exchanges1. e-Mail X X X2. Teleconference X X X3. EDI (data interchange) X X X4. EFT (fund transfer) X X X5. Product design X X X6. Distance learning X X X7. On-line information search X X X8. Automated teller machine (ATM) X X X9. Supermarket self-checkout X

10. Virtual reality for training X * 11. Virtual reality for design X * 12. GRID based teamwork *

Collaborative robotics/human-robot teams13. Telerobotic nuclear reactor repair X 14. Telerobotic mining X 15. Networked test and inspection X X X16. Robotic space exploration X * 17. Robotic construction X 18. Tele-assembly in clean room X 19. Diagnostics by sensor web X 20. Telesurgery X 21. Medical imaging and diagnostics X * 22. Robotic laser drilling X 23. Robotics in home healthcare X * 24. Virtual reality with haptic devices *

*Emerging

684 S. Y. Nof


5/24

Engineering, and Management Sciences, to address themyriad of interaction and integration challenges.

One of the main new concerns has been InformationOverload. It means that humans, each with a singlebrain,have to interact and contend with computer-basedsystems and machines that may have certain superior

skills not only in physical tasks, but also in informationhandling and computing. (Usually, a computer-basedsystem has been developed by many humans andcontains accumulation of vast knowledge). The objectivehas been to reduce the information overload (only forthe humans) by better design of interfaces and of computer systems, and by more training for the humanusers/participants.

Another natural solution has been the increasingreliance on Teamwork , not only in physical tasks, butincreasingly in design and engineering, planning, control,supply, and management tasks. Concurrent design,concurrent engineering, total information systems, andtotal quality management are examples of the urge toteam and collaborate to solve an increasingly complexworld of production and service. Further teaming hasbeen evident in the trend of outsourcing. Outsourcingcertain tasks of production and services is, practically,teaming with external enterprises. The models of supplychains and supply networks reect further teamingamong enterprises. Obviously, the enabling technologiesof computer communication increased the human abilityto share and exchange information:

. among team members;

. among distributed machines and robots, includingtele-operations and sensors; and

. among producers, suppliers, distributors andcustomers.

But at the same time, computer communication hasincreased the complexity and challenges for humanworkers and for the designers of work-systems.

The advent of the Information Age has enabled newlevels of work augmentation, including telecommunica-tion, computer Internetworking, and mobile computing.With these new powerful dimensions of work abilities,the area of e-Work has emerged. Telecommunicationand the Internet signicantly impact the cost and timeof e-activities, in certain ways potentially increasingthem, in other ways potentially reducing them. Forexample:

. The cost and the time to execute information-based transactions, per transaction, (e.g. designchange, production order, purchase order, invoicing,training, inspection, etc.) are reduced. On theother hand, the volume of transactions to correcterrors, overcome conicts, and allow for needed

customization changes and responses may increasethe overall cost of transactions.

. The cost of agency (all the middle-people) and theneeded layers of agency are reduced. On the otherhand, the lack of agency services requires moreefforts of coordination and synchronization by the

remaining layers of work.. Outsourcing becomes attractive by delegating tasks

to those who specialize in the outsourced work andcan reduce costs and delivery time. At the sametime, the cost and time of performance may actuallyincrease because of the added needs for coordinationand communication among the distributed parties.

The signicant reductions in information transfer-costand response-time in the Information Age explain theproliferation of distributed networks of producers, sup-pliers, consumers and customers. At the same time,there is further growth, to the point of explosion, in

Information Overload and Task Overload acrosslarger markets and wider distributed networks of opera-tions. The development and emergence of solutions tothese challenges are described in the next section.

3. The four circles of the e- in e-Work

An overview of the recent related research about thee- in e-Work is shown in gure 3. The four circles reectfour clusters of knowledge. Signicant amount of suchknowledge relevant to e-Work has appeared in the litera-ture; the representative sample of references in gure 3include work by members of the IFAC (InternationalFederation for Automatic Control) Committees onManufacturing and Logistics Systems (for which theauthor is the current Chair), by the PRISM Lab atPurdue, and leading work by others, with an annotatedreference list at the end of the article. A brief descriptionand some future challenges of each of the 15 subareasof e-Work appear in the following paragraphs. It isintended that this presentation will stimulate and inspirefruitful future investigations, revolutionizing the way thatbenecial e-Work can be designed.

A. e-Work

A1. Theory and models

e-Work is a collaborative, computer-supported, andcommunication enabled operation in highly distributedorganizations of humans/ robots/ autonomous systems.A goal of e-Work is to augment human abilitiesto work. Information technologies will yield benets



6/24

through e-Work if they are constructed based on proventheories and design principles.

Currently, e-Work enabling technologies in compu-ting and communication are rapidly being developed.Planning how to harness these technologies to enhancethe users ability to work and enjoy the quality of resultswill depend on layers of integrated services and support(HCI clearly being one of them). e-Work theory andmodels help to dene and match the right e-Work systemand integrate it to improve or optimize operations anddecisions.

Future research on e-Work will focus further onthe realization of the theory, and models for specicapplications.

A2. Agents

The role of agents (with work protocols and activemiddleware, which are described below) is to enable effi-cient information exchanges at the work applicationlevel, and perform tasks, under protocol control, toensure smooth, efficient interactions, collaboration, andcommunication, which augment the natural human work

abilities. An agent has been dened as a computinghardware and/or software system, that is able to (non-

autonomously, or autonomously) interact with otheragents and its environment, act, and respond reexivelyto external impacts in accordance with its given goals.

Intelligent agents (knowbots, softbots, and robots)are software and hardware entities that perform aset of tasks on behalf of a user with some degree of autonomy. They are a useful computational paradigmfor creating systems that are exible, adapt to changesof the environment, and are able to integrate heteroge-neous components and tasks. They nd applications in avariety of domains including: Internet-based informa-tion systems, adaptive (customizable) software systems,autonomous mobile and immobile robots, data miningand knowledge discovery, smart systems, decision sup-port systems, and intelligent design and manufac-turing systems. Recent research on intelligent agentsand multiagent systems builds on developments inseveral areas of computer science, including: articialintelligence (especially agent architectures, machinelearning, planning, distributed problem solving), infor-mation retrieval, database and knowledge-based systems,and distributed computing and problem solving.

Distributed control systems:Gong and Hsieh (1997)Lopez et al. (1997)Kempf et al. (2000)Esfarjani and Nof (1998)

Collaborative problem-solving:Hassard and Forrester (1997)Huang et al. (2000)Lara et al. (2000)Lara and Nof (2003)

e-Work

DistributedDecisionSupport

Integration,CoordinationCollaboration

ActiveMiddleware

Theory and models:Li and Williams (2000)Mckay and Buzzacott (2000)Nof (2000)Anussornnitisarn (2003)Ryu et al. (2003)

Agents:Jenning and Woolridge (1998)Nof (2000)Huang and Nof (2000)Villa (2002)Bellocci et al. (2003) Protocols:

Weiss (1999)Satapathy and Kumara (2000)Kim and Nof (2001)Anussornnitisarn et al. (2002)Williams et al. (2002)

Workflow:Georgakopoulos et al. (1996)Kim and Nof (1998)Leymann and Roller (2000)Kim et al. (2002)Park (2003)

Decision models:Papastavrou and Nof (1992, 1996)Khanna et al. (1998)Doumeingts et al. (2000)Ceroni and Nof (2001)

Collaboration and interaction models:Nof (1994)Rajan and Nof (1996)Montreuil et al. (2000)Chen and Nof (2001, 2003)Castro-Lacouture and Skibniewski (2003)

CIM:Barad and Nof (1997)Mejabi and Singh (1997)Banerjee and Banerjee (2000)Bullinger et al. (2000)Bubnicki (2002)

Extended enterprises:Browne et al . (1996)Wuwongse (1997)Wortmann et al. (2000)Chen (2002)Chang et al. (2003)

Humancomputer interaction:Nof (1989, 1999)Adlemo et al. (1997)Mital (1997)Duffy and Salvendy (2000)

Middleware technology:Lenart and Nof (1997)Luo et al. (1997)Huang (2002)Park (2002)Leem et al. (2003)

Distributed information systems:Etzion et al. (1995)Gardin-Goncalves et al. (1997)Kim and Nof (1998)Fakhouri et al. (2000)Jeong and Leon (2002)

GRID computing:Allcock et al. (2002)Chao et al. (2002)Foster et al. (2002)Pearlman et al. (2002)Pu et al. (2003)

Knowledge based systems:Brown et al. (1997)

Minget al.

(1997)Anumolu and Shewchuk (2000)Shakshuki et al. (2003)

Figure 3. e-Work: foundations, tools, and emerging discoveries.

686 S. Y. Nof


7/24


8/24

of information exchange. Current research focuses ondeveloping computer-supported collaborative worktools, and the model of information required by distrib-uted team members and exchange of information amongthem.

Future research will develop decision-making tools

by distributed and autonomous team members whohave different priorities and objectives, and formu-lation of resolutions (agreements) among distributedcollaborators.

B2. Humancomputer interaction (HCI)

Humancomputer interaction is an interdisciplinarysubject, which involves areas such as computer science,engineering, education, psychology and informatics. Thepurpose of HCI is to develop computer systems thatsupport people in their roles as learners, explorers, andworkers. A current research focus is on developingadvanced user interfaces that help users to efficientlyinteract with and through computer-based application

systems. Another area is the study of usability, whichmeasures the level of human ability to benet fromobjects and systems.

Future challenges include humancomputer interac-tion not only to explore the installed body of information,but also to discover new knowledge via intensive infor-

mation exchange, similar to the brainstorming processamong humans.

B3. Computer integrated manufacturing (CIM)

Computer Integrated Manufacturing research hasfocused on developing the eld of automated manufac-turing and material handling, and the application of computing to design, machine, and manufactureproducts, including quality management and processcontrol (see example in gure 5). Recent research onCIM, has involved the development and managementof the entire product lifecycle by blending severaltechnologies, including design and manufacturing;process planning; manufacturability analysis; production

LANServer

CRT

Form Board(Workstation)

CRT

CRT

TestingCenter

CRT

CRT

CRT

HUB

TestingCenter

LANServer

CRT

Form Board(Workstation)

CRT

CRT

TestingCenter

CRT

CRT

CRT

HUB

TestingCenter

Figure 5. From e-Work inside assembly facilities with TestLAN, to e-Work across supply networks and emerging supplygrids (Nof 2003).

688 S. Y. Nof


9/24

planning and scheduling; manufacturing system design;and system integration.

There are two emerging research areas for CIM:(1) Distributed support tools that govern the informationow among distributed manufacturing units that mustact under time, budget, manufacturing and environmen-

tal constraints, (2) The infusion of MEMS and nanosensors to collect and to handle information for moreintelligent real-time control and decision making inactive process control systems.

B4. Extended enterprise

Extended enterprise research is another outcome of advances in information technology. Powerful communi-cation of information enables the integration of enterprisebusiness processes beyond a single enterprise, includingfunctions and software applications from end-to-end, notonly internal but also for entire supply chains/networks(see Figure 5). Currently, the focus of research is onmodels and methodology to deliver real-time informationto enable collaboration with customers, employees andbusiness partners in a connected economy, with fasterresponse and better decisions.

In the future, research will have to address alsothe area of real-time decision making, execution, andconict management among collaborators across thesupply network.

C. Distributed decision support (DSS)

C1. Decision models

The value of decisions that are based on the decision-theoretic approach depends on the quality of availableinformation and the underlying models. Computerassistance in modelling can signicantly reduce themodel construction time, while increasing the qualityof the model, and can contribute to a wider applica-

bility of decision theory in on-line decision supportsystems. Current research focuses on the development of general-purpose knowledge representation languageswith the normative status and well-understood com-putational properties of decision-modelling formalismsand algorithms.

Future research will focus further on interaction-baseddecision models with distributed knowledge sources, andhuman interactions, which would allow the decisionstructures to be adaptive and more responsive.

C2. Distributed control system (DCS)

Distributed control systems allow the parties associatedwith a system to negotiate for their tasks and theirbenets. Autonomy is a key characteristic to classify thetype of DCS. Current research focuses on systems with

low level of autonomy among partners, such as job shopscheduling, where global objective functions are usuallygiven.

Future research will focus on systems with higher levelof autonomy, such as networks for supply chains, whereeach enterprise continues to hold its own goal andagenda, but can compromise to achieve faster satisfactorycommon goals. Similarly, in distributed robotic teams,cooperation requirement planning will be integratedwith collaborative execution of tasks (gure 6).

C3. Collaborative problem-solving

The collaborative problem-solving approach appliesinformation exchange to inuence other parties aboutcollaborators local decisions, to act coherently and toachieve a better result for the entire organization(system). Current research focuses on the developmentof methodologies and quality of solutions to mostly pre-dened problems. For example, several initiatives havebeen developed to address the needs of predenedproblems of training personnel in the armed forces.The Field Artillery School has examined the type of

Figure 6. An e-Work framework for robotic assembly as anexample of integrated support of cooperation planning; errordiagnostics, prevention and recovery; and conict diagnostics,

prevention, and resolution (Chen and Nof 2003).



10/24

collaboration necessary between military decision makersand staff units by using information handling decision-support systems in a battleeld setting. The US ArmyArmor School has investigated how students can collab-orate from remote sites to role-play with several staff positions in a virtual tactics operations center.

Future research will have to open up the possibility of collaboration among loosely-coupled entities, where lessmutual benets are acceptable, or even required amongthe collaborators.

D. Active middleware

D1. Technology

Middleware constitutes a set of services that aim at

facilitating the development of distributed applicationsin heterogeneous environments. In a way, middlewarerepresents the human functions of middle people (seesection 2). The primary objective of middleware tech-nology is to foster application portability and distributedapplication component interoperability. Middlewareoperates at a layer below the application and abovethe operating system and network layer. Commonenabling technologies include CORBA, DCOM, JavaRMI, MQSeries, and MSMQ. Recent research focusin middleware technologies is on the development of middleware architecture to provide active services toclients of the system. Similar to active protocols, the active

middleware adds intelligence to the behaviour of e-Workactivities and interactions.In the future, the active middleware service will target

more the very large scale computing platform which hascharacteristics of highly heterogeneous, autonomous, anddistributed components, beyond the typical enterprisenetwork.

D2. GRID computing

Grid computing (often shown as the term GRID)

allows access to geographically distributed computingresources. It enables sharing, selection, and aggregationof a wide variety of widely dispersed computing nodes,such as supercomputers, computing clusters, storage sys-tems, data sources, instruments, and people as a single,unied resource for solving large scale and data intensivecomputing applications. Current research focuses on thecomputational economy, on the development of Gridcomputing architecture and its components, such as ascheduler (see gure 7).

The future of this area will be the developmentof service applications which will explore the potentialof the Grid computing paradigm in both scientic andbusiness applications.

D3. Distributed information system (DIS)

The merging of computers and communications withsignicant advances in networking infrastructure hasmade accessible millions of information sources withwide varieties of information. The goal of current researchis to develop techniques to design distributed informationsystems, coordinate the distributed computation, providesecurity methods, and maintain high reliability and integ-rity of the overall information network. Corporate mem-ory systems, e-Learning, and e-Training are examples of DIS applications. The current focus is on the developmentof system architecture, standards, and methodologiesto maintain high quality of service for the large scale of distributed information networks, e.g. search robot, XMLdata management, and semantic web.

In future research, distributed information systems willbe more interactive with users in order to acquire andextract knowledge as it develops in its distributed informa-tion environment, and to serve the complex needs of users.A major challenge in DIS is the analysis and control of ripple effects when one or several nodes in the distributednetwork fail. For instance, security methods in DIS mustdetect and provide the means to handle and prevent anydamages that could be caused by intrusion at any point of the system, to assure total system performance.

D4. Knowledge-based system (KBS)

Knowledge-based systems focus on the use of knowledge-based techniques to support human decisionmaking, planning, learning, and action. Such systemsare capable of cooperating with human users. The qual-ity of support given and the manner of its presentationare important issues. Current research focuses on theconstruction and extraction of knowledge from and fora variety of information resources.

Future research will focus more on integrating KBSwith other technologies to assist a human user or agroup of users to make a better decision, especially inlarge scale distributed information systems, and forcomplex unstructured problems involving high levels of uncertainty.

The four-circle context of e-Work, discussed in thissection, can be summarized along 15 e-dimensions, asshown in table 3. These e-dimensions seem appropriatefor the design of an effective, modern work system. Four

690 S. Y. Nof


11/24

challenges for distributed e-Work design principles canbe summarized as follows:

(1) Modelling and optimization of coordinated inter-actions among autonomous entities for smoothtransitions towards a common goal;

(2) managing and controlling the growth arbitrarinessfor optimal viability and scalability;

(3) information assurance: security, integrity, andsignicance of the information which fuelse-Work, and

(4) handling errors and conicts to enable the abovethree without major disruptions.

The next section describes some of the most recentresearch and discoveries attempting to respond to thesechallenges.

4. PRISM Center developments of e-Work designprinciples

Duringresearch projects forover a decade, it hasbecomeclear that e-Work represents a new, exciting and chal-lenging model of work systems. With an ever increasing

scope of distributed and decentralized work activities,increasing distribution and magnitude of world-widemarkets, e-Work brings opportunities for better workmethods, outcome and yield by augmenting humanphysical, cognitive, temporal, and locational abilities towork. However, it has also been evident that e-Workis associated with new complex needs and new designchallenges. Table 4 summarizes a sample of projectsattempting to understand the new requirements and chal-lenges, and develop useful theories and solutions for them.

Supplyplanner

Performancemonitor

Costoptimizer

VRtutor

matcher

GriTeam Functions

Archive share Collaboration Scientificresearch

e-Learning

The NMI middleware platform

GriTeam Services and Toolkit

Resource Broker

Grid components Grid components Grid components

Examples: M.E.N optimization, scalability; MEMS sensor arrays/networks; GriTeam; FDL-CDR

DesignDesign

SupplySupply

LogisticsLogistics

ProductionProduction

DesignDesign

SupplySupply

LogisticsLogistics

ProductionProductionEnterprise i

1. Comm.

2.Exchange

3 . Integrate

Coordina-tion

Enterprise jEnterprise j

Level

ATC, IN21stC, AAE+CS+IE+ME

Tokyo--OsakaItalyS.A.PU+Chile+

Japan

P U

, U I C

, A N

L ,

I B M GM, Tecnomatix, Adept

Enterprise k Enterprise k

P R IS M L a b / P u r d u e

Figure 7. Four recent examples of e-Work research and the participating organizations and sponsors. The scope illustrated includes(clockwise from top-left) multi-enterprise network design; micro sensor arrays along arteries; co-design with conict resolution support;

and the emerging Grid-based teamwork (GriTeam). (Participants acronyms: S.A. companies in South America; PU Purdue

University; UIC University of Illinois, Chicago; ANL Argonne National Lab; IBM IBM Corp.; ATC ATC, Inc.; IN21stC Indiana 21st Century Fund for Science and Technology; AAE School of Aeronautical and Astronautical Engineering; CS

Department of Computer Science; IE School of Industrial Engineering; ME Mechanical Engineering, all at Purdue University;GM General Motors Corp.; Tecnomatix Tecnomatix Technologies, Inc.; Adept Adept Technologies, Inc.)



12/24


13/24


14/24

As shown in the table 4, dimensions of e-Work appearingin table 3 have been addressed for various applicationareas in manufacturing, production, and service. Theseefforts have been inspired by role models from success-ful, proven models of distributed and parallel computing.The theories, models, and tools that have been developedand implemented are described in detail in thegiven references. Each of the projects has resulted withrecommended design principles for effective e-Work (Nof 2002). Several key design principles, based on observa-tions about the new needs and solution approaches, aresummarized below.

8.1. The principle of cooperation requirement planning, CRP (by Rajan and Nof)

One of the most powerful augmentations of work abil-ities by e-Work is in the area of collaboration. Coopera-tion and collaboration vary from minimal information

sharing and exchange, to fully collaborative enterprises(see denitions of cooperation, collaboration, interaction,and integration modes in Nof 1994). Usually, collabora-tive e-Work implies that the parties maintain collabora-tive autonomy, but must follow, jointly, certain commonrules and full certain service agreements (contracts). Themain IT mechanisms to enable cooperation and colla-boration have been identied earlier in this article. Theyinclude agents, protocols, workow and middleware,among others. Active protocols and active middleware,with agents that can perform their tasks autonomouslyand with local initiative, are preferred. But effective col-laboration requires advanced planning (Rajan and Nof 1996a, b). Examples of planning issues are: How, when,and to whom can tasks be allocated and assigned toaccomplish the given systems goals, subject to ow andsynchronization constraints? How can pre-assigned tasksbe re-assigned to solve real-time delays or errors? Whatlevel of agent autonomy is desired? When is a local initia-tive useful, and what, if any, central authorization isrequired?

Table 4. PRISM Center development of e-Work theories, models, and tools (examples).

e-Work design feature Role model Production application Tool/model Ref.

. Workow and coordination Distributed database Computer integrated mfg. . DAFNet & AIMIS Kim (1996)Data ow Aerospace mfg. Sauvaire et al. (1998)Object orientation Laser machining cell Lenart (1993)

. Information sharingand collaboration

Task graphs Distributed designers . IDM Khanna et al. (1998)Network computing Distributed teams . Co-X Tools Eberts and Nof (1995)Internet/intranet Supply networks . Shared Process Park (2002)Internet Construction supply chain . Rebar Portal Castro (2003)Internet/intranet Production and sales . T-C-M Matsui et al. (2003)

. Collaboration planning Resource planning Multirobotic assembly . CRP Rajan (1993)

. Agent design Agent theory Manufacturing operations . ABMS Huang (1995)

. Collaboration protocoldesign

Telecom. Protocols ERP applications . TIE/P Anussornnitisarn (2003)Adaptive control Electronics testing . TestLAN/a Williams et al. (2002)

. Middleware protocols Client-servers Automotive electronicsAssembly

. RAP

. TOPEsfarjani (1994)Peralta (1996)

. Parallelism Parallel computing Global shoe design & mfg. . DPIEM Ceroni (1999)

. Resource and task allocation Local area networks Electronics assembly & test . TestLAN Williams (1994, 1999)Internet Global automotive supply . M.E.N. Method Chen (2002)

. Synchronize / re-synchronize Agent theory Robotic maintenance . ServSim Auer (2000)

. Information assurance Total quality Agent-based mfg. . (MERP) Bellocci (2001)

. Fault tolerant integration Sensor fusion Flow MEMS sensors . FTTP Liu (2001)

. Error recovery Computer recovery Robotic assembly . NEFUSER Avila-Soria (1999)

. Conict resolution Telecom Facility design by a team . FDL/CR Lara (1999)

. Multi-enterprise optimization Network ow Distributed enterprises . M.E.N. Opt. Chen (2002)

. e-learning Learning theory ERP applications . (MERP) Chen et al. (2001)

. Organizational learning Enterprise computing Manufacturing/assembly corp. . CMS Prytz et al. (1995)

. Viability measures Articial life Humanrobot facilities . TIE/A Huang (1999)

. e-work scalabili ty Distributed computer Supply networks . M.E.N. Model Chen (2002)

ABMS: Agent Based Management System; AIMIS: Agent Interaction Management System; CMS: Corporate Memory System; Co-X: CollaborativeTool for Function X; CRP: Collaboration Requirements Planning; DAFNet: Data Activity Flow Network; DPIEM: Distributed Parallel IntegrationEvaluation Method; FDL/CR: Facility Design Language/Conict Resolution; FTTP: Fault-Tolerance Time-Out Protocol; IDM: Iterative DesignModel; M.E.N.: Multi-Enterprise Network; MERP: ERP e-Learning by MBE Simulations; NEFUSER: Neuro-Fuzzy Systems for Error Recovery;RAP: Resource Allocation Protocol; ServSim: Maintenance Service simulator; TestLAN: Testers Local Area Network; TIE/A: Teamwork IntegrationEvaluator/Agent; TIE/P: Teamwork Integration Evaluator/Protocol; T-C-M: Two-Center Model; TOP: Time-Out Protocol.

694 S. Y. Nof


15/24

The Principle of Cooperation Requirement Planning, CRP,includes two phases (Rajan and Nof 1996a, b). In the rstphase CRP-I, a plan of who does what, how, and when,is generated based on the work objectives and availablefacility resources. In the second Phase, CRP-II, duringexecution, the plan is revised in real time, adapted

to temporal and spatial changes and constraints. Theadaptation responds to changes and new constraintsboth in the internal facility and its components, and inthe external design, logistics and market interactions. Arecent effort in robotic assembly and disassembly hasbeen to integrate CRP with error diagnostics, recovery,and conict resolution, as shown in gure 6 (Nof andChen 2003). The principle indicates that effective e-Work requires both advanced and adaptive, real-timeplanning of the necessary cooperation and collaboration.

8.1.1. Robotics and e-Work

It is interesting to note the link between roboticsand e-Work. Both terms originate from work. Theintelligence level of both relies on information and com-munication technologies. Robotics is a subset of e-Work,in the sense that robotic devices, including intelligentsensors, can work completely autonomously. An overalltrend in smart robotic teams normal size, micro, andnano robots is their ability to interact (by programmedand optimal control) even better than human teams. Adesign challenge in robotic teams is to achieve optimizedcoordination of e-Work interactions as a key to effective-

ness and competitiveness. CRP approaches will need to

be extended to enable the design of solutions for thecooperation and collaboration expectations.

8.2. The principle of e-Work parallelism (by Ceroni and Nof)

Every model of e-Work implies interactions amongsoftware work-spaces and human work-spaces. The prin-ciple of e-Work parallelism(Ceroni and Nof 1997, 1999,2001, 2003) is concerned with how to optimally exploitthe fact that work in these respective spaces can, andmust be allowed to advance in parallel (see table 5). Inother words, to be effective e-Work systems cannot beconstrained by linear (sequential) precedence of tasks.For example, tasks can be performed in parallel bysoftware agents preparing information for humandecision makers, while the latter are busy with othertasks, away, or asleep.

Scheduling, sequencing, and ordering work activities inparallel has long been known as a fundamental methodsimprovement principle. In e-Work, workow parallelismhas a deeper meaning, since work activities can be widelydistributed, locationally and over human- and software-spaces; there can be widely (and in effective e-Workalso wisely) distributed humanhuman interactions;humanmachine and humancomputer interactions; andmachinemachine, computercomputer interactions.The analogy between distributed computing, ande-Work in parallel interactions and networked organiza-tions of teams and facilities, is depicted in table 5. This

analogy has served well to understand the similarity and

Table 5. Analogy of e-Work with parallel and distributed computing. (following Khanna et al.1998; Anussornnitisarn and Nof 2000).

Analogy issue Parallel / distributed / GRID computing e-Work organization

Cases (examples) nCube Sensor arraySGI Origin 2000 Mobile robot team

Business enterpriseComputer GRID Supply network/grid

Challenges Routing Resource allocationData transfer Information exchangeTask allocation Task sharingInformation security Information assurance

Atomic unit CPU node Intelligent robot or sensorInternet-designer teamERP and CRM users

Communicationinfrastructure

Coupled or decoupled Communication bus;telecom. network;Internet/intranet/extranet

Communication processors

Interaction Parallel computer O.S. Competitive, cooperative,or collaborativeGRID/Network O.S.

Models Graph models; Network simulation;Message passing interface (MPI) Parallel/distributed simulation



16/24

differences, and has led to the development of the TIE/Xsimulators, which are described later in this section.

8.2.1 . KISS: Keep It Simple System!

The essence of the KISS is that the computerand communication support system can be designed ascomplex as necessary, as long as it can work autono-mously, in parallel to and supportive of humans, subjectto their inputs and instructions. The principle of e-Workparallelism identies that to be effective, e-Work systemscannot be constrained by linear (sequential) precedence of tasks, unless they are dictated by dependency constraints.KISS is associated with the principle of e-Work parallel-ism by the notion that new versions and advanced func-tions of software systems must minimize the need forrepeated human retraining, by using internal, parallele-Learning functions to simplify the transition to theevolving systems interfaces and logic.

8.2.2 . Change of work logic

Just as computer software programs, e.g. sorting, inorder to be effective do not and should not mirror thelogic of humans (although they are clearly the product of human software developers) so can e-Work not replicatetraditional human work procedures. The objective isclearly to augment human work abilities by new typesof nonlinear (nonsequential) work models; attempt toexecute in parallel whatever tasks that can beexecuted without violating dependency conditions. Asan example, consider the assembly-and-test model,TestLAN (gure 5). Instead of having assembly workers(workstations) retain exclusive control of a sharedtester while repairing diagnosed faults, a well-designtask administration protocol with access to current infor-mation about the assembly design, repair probabilitiesand time distributions, can optimize the tester sharingamong multiple, networked workstations.

8.2.3 . Multiple heads

Major challenges of e-Work stem from the power of modern IT: The information overload becomes largerand the rate at which it is reaching us is faster (highervelocity). But the individual human brain cannot (yet)cope with the ever increasing overload and communica-tion rates. Many humans put their heads together todevelop management and operational informationsystems (such as ERP, CRM, etc.), and increasingly, theexpectation that a single person must contend with those

information systems during e-Work is unreasonable, orat least requires intensive training (Huang and Nof 1998).

8.2.4 . Distributed planning of integrated executionmethod, DPIEM

DPIEM is a method for planning how to satisfy aCRP-based plan (Ceroni 1996; 1999). A key issue relatedto such planning under the e-Work parallelism principleis the ability to determine the optimal Degree Of Parallelism (DOP). DOP is dened as the maximumlevel of resource and task parallelism, which will balancethe trade-off between increased communication,transportation, and equipment costs incurred with ahigher DOP, and the increased yield (productivity)gained by that level of parallelism.

In addition to the issue of DOP, there are several basicissues in the design and implementation of any coordi-nated problem-solving and decision support system fore-Work, which can also be considered as key issues inactive middleware development. The principle of e-Work parallelism includes ve guidelines, seeking tolet the e-Work Support System (EWSS):

(1) Formulate, decompose, and allocate problems;synthesize results among groups of independententities (resources, processors) and the humansoperating and managing them.

(2) Enable applications to communicate and inter-act; the design issue is to determine under whichtask administration protocols, what, and when to

communicate during the e-Work.(3) Trigger and resynchronize independent entities to

act coherently in making decisions and takingaction; the design issue is to determine which deci-sions and actions can be assigned and delegated tohumans or to machines, and to whichhumans andmachines.

(4) Enable entities to reason about actions, plans, andknowledge of other agents (humans andmachines), and coordinate with them.

(5) Develop conict resolution, error recovery, diag-nostics and prevention: recognize and reconciledisparate viewpoints and conicting intentionsamong independent entities. (This issue is dis-cussed further under Principle 3 below.)

For instance, PIEM (with centralized optimizationalgorithms) and DPIEM (optimization with distributedprotocols) were developed for evaluating andplanning thecommunication and coordination trade-offs in e-Workwith parallelism in design, manufacturing, supply,and sales functions by Japanese, North American andSouth American companies; for outsourcing strategies;

696 S. Y. Nof


17/24

information systems middleware development; and forrobotic cells design. (Papastavrou and Nof 1996, Nof and Huang 1999, Auer et al . 2000, Auer and Nof 2000,Lara et al . 2000, Ceroni and Nof 2001, Nof 2000a,2003a, b.)

8.2.5 . The Co-X tools

Another example of the principle of parallelism are theCo-X tools (developed by Eberts and Nof 1993, 1995),where Co- implies coordinated, cooperative, and/orcollaborative support tools, e.g. Co-design, Co-plan,Co-storm (for collaborative, creative brain-storming),etc. The details of Co- can be designed at three levelsof interaction and collaboration:

Level 1 Interfaced sharing of information, ideas,and plans.

Level 2 Parallelism of information processing tasksand of physical tasks.Level 3 Beyond Level 2, addition of active pro-

tocols and agents to support (optimize)better effectiveness through parallelism.

The Co-X tools have been developed and implementedfor teaching and learning the benets and limitations of different team designs, and for lab experiments todemonstrate and investigate the interface design andusability of collaboration support (EWSS) in e-Work.

8.3. The principle of conict resolution (by Huang and Nof)

This principle addresses the cost of resolving conictsamong collaborating e-Workers. It has been observedand recognized that with a greater rate of interactions,which increases with the number of collaborating parties,there is also a greater occurrence of conicts and errors.This problem is critical in terms of scalability. It meansthat beyond the information and task overloads, e-Workmust be designed to overcome quickly and inexpensivelyas many errors and conicts as required to be effective.To illustrate this issue, suppose: q is the relative portion of human intervention in the resolution process; s is theprobability of resolving a conict in any one iterationof negotiations. Huang and Nof (1999a) showed thatfor practical values of s, the resolution cost would dependon q as follows:

(a) With higher q, i.e. more human interaction andinvolvement are needed, the cost of resolutiongrows exponentially and it is unbounded. Implica-tion: Under this design conditions, the e-Worksystem will collapse because of ineffectiveness.

(b) With a smaller value of q, approaching zero,meaning IT services are designed and applied forresolution-support by automating functions of negotiations for corrections or compromises, thenthe total cost of resolution reaches an upperbound. Implication: In this case, e-Work systems

can be effective.

The design objective, following this principle, is todevelop better and more powerful IT support (bringingthe value of q to a minimum) for detection, prevention,and resolution or recovery of errors and conicts, leadingto lower overall costs, better quality results, and moreeffective e-Work. Further details are explained byHuang et al. (2000); Anussornnitisarn and Nof (2003).

Conict resolution for collaborative facility design byFDL/CR was developed by Lara (1999), Lara and Nof (1999a, b, 2003) and knowledge-based diagnostics andrecovery by NEFUSER were developed by Avila-Soria

and Nof (2000) (see table 4). For example, FDL/CRencompasses rational execution of pre-ordered conictresolution approaches, from the simpler and least costly,to the more complex and costly resolution mechanisms:beginning with support for direct negotiations; if thatstage fails, moving to third-party mediation; then incor-porating additional parties; if still not resolved, movingon to persuasion; and if all else fails, applying support forarbitration. This method incorporates principles for pre-vention of conict perpetuation and escalation toimprove short-term performance, and computer-basedlearning to improve usefulness and effectiveness of futureresolution tasks. To incorporate prevention principles,the last stage of the method, arbitration, is sequentiallyorganized in three phases of conict modelling and ana-lysis:

(1) Incorporation of principles for prevention of con-ict perpetuation and escalation, to improve thedetection, resolution, and consequently, thee-Work performance;

(2) implementation of computer-based learning, toincrease the usefulness of the resolution-supportmethod; and

(3) integration of conict detection and resolution,which results in an increased effectiveness, and

potentially better quality of the facility design.

Control Protocols for the diagnostics, prevention, reso-lution, and recovery of errors and conicts are importantdesign issues. Future collaborative e-Work, includingrobotics and teamwork, will depend on cheaper, redun-dant groups, arrays, networks, and grids of interactingparties, e.g. business partners, human designer teams,humanrobot task forces for security and rescue opera-tions, and DNA-sensor arrays. Enabling effective e-Work



18/24

in each of these examples will require effective handlingof errors and conicts.

8.4. New measures and evaluation

New work models, as those illustrated in table 4 forthe design of effective e-Work, require new measures inaddition to traditional measures of effectiveness. Some of the measures have already emerged with the adventof CIM and networked enterprises: Flexibility, agility,connectivity, integration-ability, scalability, reachability,and so on. Two recent measures of particular interest toe-Work design are level of autonomy, and viability (e.g.Huang 1999).

The level of autonomy relates to the delegationof authority, decentralization, and to how active theagents, protocols, and other e-Work system participantscan be. Viability in the context of e-Work (e.g. of

the whole e-Work system, or of its components) hasbeen dened as a relative measure of the ratio betweenthe cost of operating and sustaining agents, and therewards from their services. Theories of agent design

and agent interaction models have been developedusing measures of autonomy and viability (Huang andNof 1999b,c). Recently, Anussornnitisarn (2003) hasdiscovered that e-Work rationalized by viability-based protocols performs signicantly better than byprotocols ignoring the viability measures. Examples of

other measures emerging as useful for e-Work designare learning-ability, collaboration-ability; and relatedto errors and conicts, interesting measures areprevention-ability, detect-ability, diagnosis-ability, andrecovery-ability. A critical area to e-Work is informationassurance, including measures of security, integrity, andinformation signicance (Bellocci 2001, Bellocci et al .2003).

Combinations of O.R. and A.I. models and simulationhave been developed to measure, understand, evaluate,design and plan e-Work systems. Exploiting theanalogy between e-Work systems, and parallel andnetworked computing, (see table 5), a family of parallelsimulators, TIE/X has been developed and implemented(Anussornnitisarn and Nof 2000). These simulators(gures 4 and 8) can assign autonomous CPUs to repre-

OO vvee r r vviiee ww oo f f TTIIEE dd ee vvee lloo pp mm ee nn tt

e-Work

Engineering view

Enterpriseview

Supply-chain/Networkview

TIE/ Design

TIE/ Agent

TIE/ Protocol

Design ofbusiness/Mfg/Transp.

processes

Design ofautonomous system

& structure

Design of collaborativeinteraction protocols

DAF-Net

AIMIS Adaptive coordination

TIE/MEMS

nCube parallel computer using HPF

Intel Paragon X/PSusing MPI

SGI Origin 2000using MPI

TIE/Protocol

DAF-Net & AIMISusing PVM

Adaptive protocolon TestLAN

TIE/Design

TIE/Agent

TIE/GROUP

Modeling perspective:

Implementation of simulators:

Figure 8. The Teamwork Integration Evaluator (TIE). TIEs purpose: use the analogy of parallel/ distributed/ networked computersto model e-Work environments networked designers; agents; protocols; sensor arrays; autonomous groups (Sponsors of TIE

development include NSF, IBM, Caterpillar, SGI, TAP, Tecnomatix, and Indiana 21st Century Fund for Science Technology.)Note: Tools and models mentioned in this gure are listed in table 4.

698 S. Y. Nof


19/24

sent each participating entity in an e-Work team, observetheir interactions under given task administration proto-cols, and generate performance evaluation measures tohelp the designers of alternative e-Work systems.

9. Conclusions and future challenges

The way people and systems operate, work, commu-nicate, and collaborate is being transformed by e-Work.Hence, understanding how to design effective e-Work isfundamental to productivity and competitiveness.Models of e-Work have been developed to address designissues, from interaction to coordination, conict resolu-tion and error recovery, and the management of complexe-Work production and service environments. Designand decision tools and methods have been developedfor this modelling, with the objective of understandingthe emerging requirements, limitations, and potentialcapabilities of e-Work. Modelling challenges for e-Workdesign include enhancements of models, such as TIE/X,to include further e-Work functions along the 15e-dimensions in table 3, which depend on emergingInternet, Intranet, and related technologies. Otherchallenges include the evaluation and assessment of newmeasures, mentioned in the previous section, such ascollaboration-ability, conict prevention-ability, anderror detection-ability.

Based on extensive studies of e-Work systems, discov-eries of several design principles have been discussed infour main areas:

. Cooperation requirement planning

. e-Work parallelism

. Error and conict handling

. e-Work effectiveness measures.

Effective e-Work is fundamental to e-activities, such ase-Commerce, e-Business, e-Learning, intelligent trans-portation, and e-Manufacturing. Enabling services todecouple applications from the computer and communi-cation layers can augment human work signicantly byperforming many tasks in parallel, by software and hard-ware agents, thus reducing the information overload and

task overload currently imposed on human workers. Inthe next generation of e-Work, based on evolving IT,systems will be developed with inherent collaborativesupport tools in large-scale e-Work environments.

Several open questions remain: Will people like towork in the collaborative e-environment? Will workersand managers be willing to trust the results obtainedand delivered by the agents work? Will they acceptcomputer-supported negotiations? The design of effectivee-Work will also have to address these issues.

Acknowledgement

Research reported in this article has been developed atthe PRISM Center with NSF, Indiana 21st Century fundfor Science and Technology, and industry support.Special thanks to my colleagues and students at the

PRISM Lab, who have worked with me to develop thee-Work knowledge, and to Dr. P. Anussornnitisarn whohelped me in preparing the Four Circle review.

Additional details about the PRISM Center activitiescan be found at URL http://gilbreth.ecn.purdue.edu/

prism.

References

ADLEMO , A., ANDREASSON , S.-A., G ULLANDER , P., F ABIAN , M.,and L ENNARTSON , B., 1997, Operator control activities inexible manufacturing systems. International Journal of Computer Integrated Manufacturing, 10(14), 221232.

ALLCOCK , B., BESTER , J., BRESNAHAN , J., C HERVENAK ,A. L., F OSTER , I., K ESSELMAN , C., S AM , M., V ERONIKA , N.,Q UESNEL , D. , and T UECKE , S. 2002, Data managementand transfer in high-performance computational gridenvironments. Parallel Computing, 28(5), 749771.

ANUMOLU , B., and S HEWCHUK , J. P., 2000, Design of a toolingdatabase implementation for an existing facility. Computers inIndustry, 42(23), 221229.

ANUSSORNNITISARN , P., 2003, Design of middleware protocolsfor the distributed ERP environment, PhD dissertation,School of Industrial Engineering, Purdue University, WestLafayette, IN.

ANUSSORNNITISARN , P., and N OF , S. Y., 2000, TIE protocol, ateamwork integration evaluator for coordination protocols.In Proceedings of ICPR-2000, Bangkok, Thailand, August. OnCD-Rom.

ANUSSORNNITISARN , P., and N OF , S. Y., 2003, Decentralizedcontrol of cooperative and autonomous agents for solving thedistributed resource allocation problem. In Proceedings of ICPR-17 , Blacksburg, VA, August. On CD-Rom.

ANUSSORNNITISARN , P., P ERALTA , J., and N OF , S. Y., 2002,Time-out protocol for task allocation in multi-agent system, Journal of Intelligent Manufacturing, 13(6), 511522.

AUER , J. E., 2000, Agent-based prediction of customer require-ments for distributed stream service systems, MSIE thesis,Purdue University.

AUER , J. E., and N OF , S. Y., 2000, Agent-based prediction of customer requirements for distributed steam service systems.Research Memo 20008, School of IE, Purdue University, IN.

AUER , J. E., H UANG , C. Y., and N OF , S. Y., 2000, Simulationsoftware to model distributed service systems. ResearchMemo 20005, School of IE, Purdue University, IN.

AVILA -SORIA , J., 1999, Interactive error recovery for roboticassembly using a Neural-Fuzzy approach, MSIE Thesis,School of IE, Purdue University, West Lafayette, IN.

AVILA -SORIA , J., and N OF , S. Y., 2000, Interactive error recov-ery for robotic assembly using a neural-fuzzy approach.Research Memo 20009, School of IE, Purdue University,IN.



20/24

BANERJEE , A., and B ANERJEE , P., 2000, A behavioral scenegraph for rule enforcement in interactive virtual assemblysequence planning. Computers in Industry, 42(23), 147158.

BARAD , M., and N OF , S. Y., 1997, CIM exibility measures: areview and a framework for analysis and applicability assess-ment. International Journal of Computer Integrated Manufacturing,10(14), 296308.

BELLOCCI , T., 2001, Planning variable information assurancein agent-based workow systems, MSIE thesis, PurdueUniversity, West Lafayette, IN.

BELLOCCI , T., L EHTO , M. R., and N OF , S. Y., 2003, Assuringinformation quality in industrial enterprises: experiments inan ERP environment. In D. Harris, V. Duffy, M. Smith,& C. Stephanidis (eds) Human-Centered Computing:Cognitive, Social and Ergonomic Aspects. Proceedings of the10th International Conference on HumanComputer Interaction,Crete, Greece, June 2226 (New Jersey: LaurenceErlbaum), 3, 654658.

BRADY , K., and T ARN , T.-J., 2000, Internet based manufactur-ing technology: intelligent remote teleoperation. In (eds)Proceedings of 2000 IEEE/RSJ International Conference onIntelligent Robots and Systems (IROS 2000), Takamatsu, Japan,

November 2000 (New Jersey: IEEE), pp. 843848.BROWN , A. W., J UDD , R. P., and R IDDICK , F., 1997,Architectural issues in the design and implementation of anintegrated toolkit for manufacturing engineering. International Journal of Computer Integrated Manufacturing, 10(14), 142159.

BROWNE , J. , H ARHEN , J. , and S HIVNAN , J., 1996, ProductionManagement Systems: An Integrated Perspective, 2nd Edn(Wokingham, UK: Addison-Wesley).

BUBNICKI , Z., 2002, Learning process in a class of computerintegrated manufacturing systems with parametric uncertain-ties. Journal of Intelligent Manufacturing, 13(6), 409416.

BULLINGER , H.-J., W ARSCHAT , J., and F ISCHER , D., 2000,Rapid product development an overview. Computers inIndustry, 42(23), 99108.

BURGESS , W., 2003, Report of the e-work action forum 2002,Government of Ireland, Dublin.

CASTRO -L ACOUTURE , D., 2003, B2B E-Work intranet solutiondesign for rebar supply interactions, PhD dissertation,Purdue University, West Lafayette, IN.

CASTRO -L ACOUTURE , D., and S KIBNIEWSKI , M. J., 2003,Applicability of e-Work models for the automation of con-struction materials management systems. Journal of ProductionPlanning and Control (this issue).

CERONI , J. A., 1996, A framework for trade-off analysisof parallelism in production operations, MSIE thesis,Purdue University, IN.

CERONI , J. A., 1999, Models of integration with parallelism of distributed organizations, PhD Dissertation, PurdueUniversity, IN.

CERONI , J. A., and N OF , S. Y., 1997, Planning effectiveparallelism in production operations. Research Memo 9710, School of IE, Purdue University, IN.

CERONI , J. A., and N OF , S. Y., 1999, Planning integratedenterprise production operations with parallelism. InProceedings ICPR-15 , Limerick, Ireland, August (eds) M. T.Hillery & H. J. Lewis (Ireland: University of Limerick), pp.457460.

CERONI , J. A., and N OF , S. Y., 2001, The parallel computingapproach in production system integration modeling.Management and Control of Production and Logistics (MCPL2000) Proceedings of the 2nd IFAC/IFIP/IEEE Conference,

Grenoble, France, July 2000. 3, pp. 12571262 (Kidlington,UK: Elsevier Science).

CERONI , J. A., and N OF , S. Y., 2003, A workow model basedon parallelism for distributed organizations. Journal of Intelligent Manufacturing 13(6), 439462.

CERONI , J . A. , M ATSUI , M., and N OF , S. Y., 1999, Com-munication based coordination modeling in distributed

manufacturing systems . Internatinal Journal of ProductionEconomics, 60 61, 281287.CHANG , T.-W., S HIN , K., and P ARK , J., 2003, A develop-

ment methodology for e-Work ontology using RDF/RDFSand PSL. Journal of Production Planning and Control (thisissue).

CHAO , K.-M., A NANE , R., C HEN , J.-H., and G ATWARD , R.,2002, Negotiating agents in a market-oriented grid.Proceedings of the 2nd IEEE/ACM International Symposium onCluster Computing and the Grid (CCGRID-02), Berlin,Germany, May 2002 (NJ: IEEE), 436437.

CHEN , J., 2002, Modeling and analysis of coordinationfor multi-enterprise networks, PhD dissertation, Schoolof Industrial Engineering, Purdue University, WestLafayette, IN.

CHEN , J., and N OF , S. Y., 2001, Scalable multi-enterprise net-work. PRISM Symposium and Reunion, 911 August, Schoolof Industrial Engineering, Purdue University. On CD-RomISBN 0-9316 82-92-4.

CHEN , J. , Y ERUSHALMI , M., W ARD , J. E., and N OF , S. Y.,2001, The use of MICSS/MERP, an ERP model, ineducation and research. Proceedings of IFAC Workshop onManufacturing Modeling for Management and Control , August,Prague, C.R., pp. 6671.

CHESHIRE H ENBURY UK, 2003, 27 July, http://www.cheshirehenbury.com/ebew/

CHIOZZA , E., and S TANFORD -SMITH , B., (eds), 2001, E-Workand E-Commerce: Novel Solutions and Practices for a Global Networked Economy (Amsterdam, The Netherlands: IOSPress).

DOUMEINGTS , G., D UCQ , Y., V ALLESPIR , B., and K LEINHANS ,S., 2000, Production management and enterprise modelling.Computers in Industry, 42(23), 245263.

DUFFY , V. G., and S ALVENDY , G., 2000, Concurrent engineer-ing and virtual reality for human resource planning.Computers in Industry, 42(23), 109126.

EBERTS , R. E., and N OF , S. Y., 1993, Distributed planning of collaborative production. International Journal of Advanced Manufacturing Technology, 8, 258268 (Norcross, GA: IEEE).

EBERTS , R. E., and N OF , S. Y., 1995, Tools for collaborativework. Proceedings of IERC 4, Nashville, TN, pp. 449455.

ESFARJANI , K., 1994, Planning client-server integration proto-cols for test work-cells, MSIE thesis, Purdue University, WestLafayette, IN.

ESFARJANI , K., and N OF , S. Y., 1998, A clientserver

model for integration and collaboration in productiontesting. International Journal of Production Research, 36(12),32953321.

ETZION , O. D ORI , D., and N OF , S. Y., 1995, Active coordina-tion of a CIM multi-database system. International Journal of Computer Integrated Manufacturing, 8, 116125.

FAKHOURI , S. A., J EROME , W. F., N AIL , V. K., R AINA , A., andVARMA , P., 2000, Active middleware services in a decisionsupport system for managing highly available distributedresources, Proceedings of Middleware 2000, 47 April 2000,New York (Berlin: Springer-Verlag), pp. 349371.

700 S. Y. Nof


21/24

FOSTER , I., K ESSELMAN , C., N ICK , J. M., and T UECKE , S. 2002,Grid services for distributed system integration. Computer ,35(6), 3746.

GARDIN -G ONCALVES , R., V ITAL , M., S OUSA , P., S EIGER -GARCAO , A., and P AMIES -T EIXEIRA , J., 1997, Implementa-tion of computer integrated manufacturing systems usingSIP: CIM case studies using a STEP approach. Journal of

CIM , 10(14), 172180.GEORGAKOPOULOS , D., H ORNICK , M. F., and M ANOLA , F.,1996, Customizing transaction models and mechanisms in aprogrammable environment supporting reliable workowautomation. IEEE Transactions on Knowledge & DataEngineering, 8(4), 630649.

GONG , D.-C., and H SIEH , Y.-W., 1997, Conceptual design of a shop oor control information system. International Journal of Computer Integrated Manufacturing, 10(14), 416.

H AMEL , W. R., 2000, e-Maintenance robotics in hazardousenvironments. Proceedings of 2000 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2000).Takamatsu, Japan, November 2000 (NJ: IEEE), pp. 838842.

H ASSARD , J., and F ORRESTER , P., 1997, Strategic and organi-

zational adaptation in CIM systems development.International Journal of Computer Integrated Manufacturing,10(14), 181189.

H UANG , C. Y., 1995, Formation and evaluation of agent net-works for manufacturing systems, MSIE thesis, PurdueUniversity, West Lafayette, IN.

H UANG , C. Y., 1999, Autonomy and viability in agent-basedmanufacturing systems, PhD dissertation, School of IE,Purdue University, West Lafayette, IN.

H UANG , C. Y., 2002, Distributed manufacturing executionsystems: a workow perspective. Journal of Intelligent Manufac-turing, 13(6), 485498.

H UANG , C. Y., and N OF , S. Y., 1998, Development of inte-grated models for material ow design and control a toolperspective. Robotics and CIM , 14, 441454.

H UANG , C. Y., and N OF , S. Y., 1999a, Enterprise agility: a viewfrom the PRISM lab. International Journal of Agile Management Systems, 1(1), 5159.

H UANG , C. Y., and N OF , S. Y., 1999b, Autonomy and viability measures for agent-based manufacturing systems. ResearchMemo 9922, School of IE, Purdue University, IN.

H UANG , C. Y., and N OF , S. Y., 1999c, Viability a measurefor agent-based manufacturing systems, In (eds) M. T.Hillery & H. J. Lewis, Proceedings of ICPR, Limerick,Ireland (Limerick, Ireland: University of Limerick), pp.18691872.

H UANG , C. Y., and N OF , S. Y., 2000, Formation of autonomousagent networks for manufacturing systems . International Journal of Production Research, 38(3), 607624.

H UANG , C. Y., C ERONI , J. A. and N OF , S. Y. 2000, Agility of

networked enterprises: parallelism, error recovery and con-ict resolution, Computers in Industry, 42(23), 275287.

JENNINGS , N. R., and W OODRIDGE , M. J., 1998, AgentsTechnology Foundation, Applications and Markets(New York:Springer-Verlag).

JEONG , I.-J., and L EON , V. J., 2002, Decision-making and coop-erative interaction via coupling agents in organizationallydistributed systems. IIE Transactions, 34(9), 789802.

K EMPF , K . , U ZSOY , R . , S MITH , S., and G ARY , K., 2000,Evaluation and comparison of production schedules.Computers in Industry, 42(23), 203220.

K HANNA , N. K., F ORTES , J. A. B., and N OF , S. Y., 1998, Aformalism to structure and parallelize the integration of coop-erative engineering design tasks. IIE Transactions , 30(1), 115.

K IM , C. O., 1996, DAF-net and multi-agent based integrationapproach for heterogeneous CIM information systems. PhDdissertation, Purdue University, West Lafayette, IN.

K IM , C. O., and N OF , S. Y., 1998, Coordination and integra-

tion models for distributed and heterogeneous CIMInformation. International Journal of Industrial Engineering,Applications and Practice, Special Issue on Information Systems,5(1), 6877.

K IM , C. O., and N OF , S. Y., 2000, Investigation of PVM for theemulation and simulation of a distributed CIM workowsystem. Internatinal Journal of CIM , 13(5), 401409.

K IM , C.-O., and N OF , S. Y., 2001, Design of collaborationframework for distributed CIM data activities, IIE Transactions, 33(7), 53546.

K IM , C. O., J UN J., K IM , S., and B AEK , J. K. 2002, Contract-collaboration network method for modeling manufacturingresource control workows. Journal of Intelligent Manufacturing, 13(6), 463476.

LARA , M. A. 1999, Conict resolution in collaborative facility

design, PhD dissertation, Purdue University, IN.LARA , M. A., and N OF , S. Y., 1999a, Computer-supportedconict resolution in collaborative facility design. ResearchMemo 9921.

LARA , M. A., and N OF , S. Y., 1999b, Systematic resolution of conicts situations in collaborative facility design. ResearchMemo 9923, School of IE, Purdue University, IN.

LARA , M. A. , and N OF , S. Y., 2003, Computer-supportedconict resolution for collaborative facility designers.International Journal of Production Research, 41(2), 207234.

LARA , M. A., W ITZERMAN , J. P., and N OF , S. Y., 2000, FDL.Facility description language for integrating distributiondesigns. International Journal of Production Research, 38(11),24712488.

LEEM , C. S., P ARK , N., S UH , H. S., and K IM , B. Y., 2003,

A study on e-Transformation effectiveness analysis withcases in manufacturing and retail industry. Journal of Production Planning and Control (this issue).

LENART , G. M., 1993, Object-oriented integration of concur-rent engineering and a laser processing cell. MSIE thesis,Purdue University, IN.

LENART , G. M., and N OF , S. Y., 1997, Object-oriented designof information integration in sheet metal manufacturing.International Journal of CIM , 10(14), 2950.

LEYMANN , F., and R OLLER , D., 2000, Production Workow:Concepts and Techniques (Upper Saddle River, New Jersey:Prentice Hall).

L I , H., and W ILLIAMS , T. J. 2000, The interconnect chain of enterprise as presented by the Purdue Enterprise ReferenceArchitecture. Computers in Industry, 42 (23), 265274.

L IU , Y., 2001, Distributed micro ow-sensor arrays andnetworks: design of architecture and fault-tolerantintegration. MSIE thesis, Purdue University, WestLafayette, IN.

LOPEZ , J. M., L LORENTE , J. I., B URGOS , A., and M ARTIJA , I.,1997, A software engineering method for the design of dis-crete manufacturing cell control. International Journal of Computer Integrated Manufacturing, 10(14), 126141.

LUO , J. L., W ONG , W. C. K., and Y AN , J. Q., 1997, Object-oriented modeling for CAPP open architecture. Journal of Computer Integrated Manufacturing, 10(14), 7482.



22/24

M ATSUI , M., A ITA , S., N OF , S. Y., C HEN , J., and N ISHIBORI ,Y., 2003, Analysis of cooperation effects in two-center pro-duction models. International Journal of Production Economics, 84,101112.

M CK AY , K. N., and B UZACOTT , J. A., 2000, The application of computerized production control systems in job shop envi-ronments. Computers in Industry, 42 (23), 7998.

M EJABI , O. O., and S INGH , N., 1997, A framework forenterprise-wide integration. International Journal of Computer Integrated Manufacturing, 10(14), 212220.

M ENG , M., C HENG , C., L IU , P. X., and R AO , M., 2000, E-Service robot in home healthcare. Proceedings of 2000 IEEE/RSJ International Conference on Intelligent Robots and Systems(IROS 2000) , Takematsu, Japan, November 2000 (NJ:IEEE) pp. 832837.

M ING , X. G. , M AK , K . L . , Y AN , J . Q. , M A, D. Z., andZHANG , H. Q., 1997, A hybrid expert-neural-based functionmodel for CAPP. International Journal of Computer Integrated Manufacturing, 10(14), 105116.

M ITAL , A., 1997, What role for humans in computer integratedmanufacturing?, International Journal of Computer Integrated Manufacturing, 10(14), 190199.

MONTREUIL

, B., FRAYRET

, J.-M., and DAMOURS

S., 2000, Astrategic framework for networked manufacturing, Computersin Industry, 42(23), pp. 299218.

NEW WAYS , 2003, New Ways of Living and Working,e-Working in Ireland Code of Practice, 27 July, www.entemp.ie/e-work.htm

NOF , S. Y., 1989, Knowledge-based real-time machine interac-tion. Proceedings of 22nd Hawaii International Conference on SystemSciences, Kailua-Kona, Hawaii., 3, pp. 192199 (Washington:IEEE Computer Soc. Press).

NOF , S. Y. 1994, Information and Collaboration Models of Integration(Dordrecht, NL: Kluwer Academic Publishers).

NOF , S. Y., 1995, Parallel and distributed models of integrationin production systems. In (eds) E. M. Dar-El, R. Karni, &Y. Herer. Proceedings of the ICPR-13 , Jerusalem, Israel, August(London: Freund).

NOF , S. Y., 1999, Robot ergonomics: optimizing robot work, In(ed.) S. Y. Nof, Handbook of Industrial Robotics 2nd Edition(NewYork: John Wiley & Sons), pp. 603604.

NOF , S. Y., 2000a, Models of e-Work. Proceedings of the IFAC/MIM-2000 , Rio, Greece, pp. 521527 (Hungary: IFAC).

NOF , S. Y., 2000b, Intelligent collaborative agents, In (eds) C.Moore et al . Encyclopedia of Science and Technology(New York:McGraw Hill), pp. 219222.

NOF , S. Y., 2000c, Models and tools for e-Work. In (ed.) S.Raczynski, Proceedings of the Vth International Conference on AI and Simulation, Mexico City, Mexico (Mexico City: McleodInstitute, Universidad Panamericana), pp. 249258.

NOF , S. Y., 2001, To do good e-Business, somebody has to dothe e-Work. Lecture at the e-Enterprise Seminar, PurdueUniversity, February.

Nof, S. Y., 2002, IE590E: Design of e-Work and e-Businesssystems. Course notes, Purdue University, West Lafayette IN.

NOF , S. Y., 2003a, Collaborative robotics and e-Work.Proceedings of the International Symposium on Robotics,Rosemont, IL, June 25 (Ann Arbor, MI: RoboticsIndustries Assoc.).

NOF , S. Y., 2003b, Computer-supported integration andcollaboration of distributed e-Work: review of models andtechniques. PRODLOG presentation at Prospective of Researchon Production/Logistics, lEcole des Mines de Saint-Etienne,France, 19 March.

NOF , S. Y., and C HEN , J., 2003, Assembly and disassembly: anoverview and framework for cooperation requirement plan-ning with conict resolution. Journal of Intelligent & RoboticSystems,37(3), 307320.

NOF , S. Y., and H UANG , C. Y., 1999, Enterprise agility: a viewfrom the PRISM lab. International Journal of Agile Management Systems, 1(1), 5159.

PAPASTAVROU , J. D., and N OF , S. Y., 1992, Decision integra-tion fundamentals in distributed manufacturing topologies.IIE Transactions , 24(3), 2742.

PAPASTAVROU , J. D., and N OF , S. Y., 1996, Analytic proceduresfor optimizing engineering task integration topologies.Decision Support Systems, 17(3), 159182.

PARK , N., 2002, A shared process model for independent andsynchronized e-business transactions. Journal of Intelligent Manufacturing, 13(6), 499510.

PARK , N., 2003, Collaboration and integration of business pro-cesses based on RosettaNet standard and process ow control.Production Planning and Control (this issue).

PEARLMAN , L., W ELCH , V., F OSTER , I., K ESSELMAN , C., andT UECKE , S., 2002, A community authorization service forgroup collaboration. Proceedings Third International Workshop

on Policies for Distributed Systems and Networks, 57 June 2002,Monterey, CA (Los Alamitos, CA: IEEE Computer Soc.),pp. 5059.

PERALTA , J. I., 1996, Evaluation of time-out protocols in man-ufacturing systems, MSIE thesis, Purdue University, IN.

PERALTA , J. A., A NUSSORNNIHSARN , P., and N OE , S. Y., 2002,Analysis of a time-out protocol and its applications in a singleserver environment. International Journal of CIM , 16(1), 113.

PRYTZ , K. , N OF , S. Y., and R OLSTADAS , A., 1995, Manu-facturing integration and learning by corporate memorysystems (CMS). Research Memo 9512, School of IE,Purdue University, IN.

PU , J., X IONG , Z., and W U , Z., 2003, The research onQoS for grid computing, Proceedings of ICCT 2003 International Conference on Communications Technology, 911

April 2003, Beijing, China (Beijing: Beijing Univ. Posts &Telecommun. Press), pp. 17111714.R AJAN , V. N., 1993, Cooperation requirement planning for

multi-robot assembly cells, PhD dissertation, PurdueUniversity, IN.

R AJAN , V. N., and N OF , S. Y., 1996a, Minimal precedenceconstraints for integrated assembly and execution planning .IEEE Transaction on Robotics and Automation, Special Issue onAssembly and Task Planning, 12(4), 175186.

R AJAN , V. N., and N OF , S. Y., 1996b, Cooperation requirementplanning (CRP) for multiprocessors: optimal assignment andexecution planning. Journal of Intelligent and Robotic Systems, 15,419435.

R OVETTE , A., 2000, Telerobotic surgery control and safety.2000 ICRA. IEEE International Conference on Robotics and

Automation. 2428 April 2000, San Francisco, USA(Piscataway, NJ: IEEE) .

R YU , K., S ON , Y., and J UNG , M., 2003, Framework for fractal-based supply chain management of e-Work company. Journal of Production Planning and Control (this issue).

SATAPATHY , G., and K UMARA , S. R. T., 2000, Negotiation fortransportation tasks with stochastic payoffs. Computers inIndustry, 42(23), 193202.

SAUVAIRE , P., C ERONI , J. A., and N OF , S. Y., 1998,Information management for FMS and nonFMS decisionsupport integration. International Journal of Industrial

702 S. Y. Nof


23/24

Engineering Application and Practice, Special Issue onInformation Systems, 5(1), 7887.

SHAKSHUKI , E., G HENNIWA , H., and K AMEL , M., 2003, Anarchitecture for cooperative information systems. Knowledge-Based Systems, 16(1), 1727.

V ILLA , A., 2002, Autonomy versus efficiency in multi-agentmanagement of extended enterprises. Journal of Intelligent

Manufacturing, 13(6), 429438.WEISS , G., 1999, Multiagent Systems: a Modern Approach toDistributed Articial Intelligence(Cambridge, MA: MIT Press).

W ILLIAMS , N. P. 1994, The TestLAN approach to the designof testing systems, MSIE thesis, Purdue University, IN.

W ILLIAMS , N. P., 1999, The effectiveness of protocol adaptabil-ity in TestLan production environments, PhD dissertation,Purdue University, IN.

W ILLIAMS , N. P., and N OF , S. Y., 1998, User manual for theTestLAN simulator. Research Memo 986, School of IE,Purdue University, IN.

W ILLIAMS , N. P., and N OF , S. Y., 1999, Protocol adaptabilityin TestLAN-based production environments. ResearchMemo 991, School of IE, Purdue University.

W ILLIAMS , N. P., L IU , Y., and N OF , S. Y., 2002, The TestLANapproach and protocols for the integration of distributedassembly and test networks. International Journal of ProductionResearch, 40(17), 45054522.

W ILLIAMS , N. P., L IU , Y., and N OF , S. Y., 2003, Analysis of workow protocol adaptability in TestLAN productionsystems. IIE Transactions on Design and Manufacturing, 35,965972.

W ORTMANN , J. C., H EGGE , H. M. H., and R OLEFES , S., 2000,Embedding enterprise software in extended enterprisemodels. Computers in Industry, 42(23), pp. 231244.

W UWONGSE , V., 1997, A deductive object-oriented approach toCIM integration. International Journal of Computer Integrated Manufacturing, 10(14), 6173.



24/24

Documents

Design of effective ework