Operations Research Management Science Today - June 2001 - Decision Support Systems

<BODY bgcolor="#ffffff" link="cc0000" vlink="cc0000" text="#000000"> <img src="../cleardot.gif" width=1 height=1 vspace=4 border=0><br>   <table border=0 cellspacing=0 cellpadding=0 width=480><tr valign=top> <td width=10><img src="../cleardot.gif" width=1 height=1 hspace=4 border=0></td> <td width=470> <FONT face="verdana, arial, helvetica" SIZE=2><B><I><A HREF="../../ORMS.shtml" TARGET="_top">OR/MS Today</A> - June 2001</I></B></font><BR> <hr size="1"> <img src="cleardot.gif" width=1 height=1 vspace=4 border=0><br> <font face="times new roman, times"> <img src="../cleardot.gif" width=1 height=1 vspace=4 border=0><br> <FONT face="verdana, arial, helvetica" SIZE=2><B>Decision Support Systems</B></FONT><br> <IMG SRC="../barbluehook.gif" WIDTH="468" HEIGHT="4"><br> <img src="../cleardot.gif" width=1 height=1 vspace=8 border=0><br> <FONT SIZE="+2"><B>Empower to the People</B></FONT><BR> <img src="../cleardot.gif" width=1 height=1 vspace=2 border=0><br> <FONT SIZE="+1"><B><I>How decision support systems and IT can aid the OR analyst</I></B></FONT><BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> <FONT face="verdana, arial, helvetica" SIZE=2><B>By Les Foulds and Cherian Thachenkary</B></FONT><br> <hr size="1"> <img src="../cleardot.gif" width=1 height=1 vspace=12 border=0><br>   The recent explosion of information technology has made the practice of OR much more widespread than in previous decades. In the past, in order to address complex problems arising in industry, many OR practitioners relied on constructing large-scale models and attempted to solve them using optimization techniques such as mathematical programming (MP). A growing body of opinion questions whether such an approach, on its own, can deal successfully with many of the complex, ill-defined, difficult-to-model issues now facing OR practitioners. This has given rise to other approaches, such as soft systems methodology, to tackle what Ackoff termed today's "messes."<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> The power of MP can be enhanced by incorporating its models and methods within a decision support system; taking advantage of modern information technology. Such a system, containing MP subroutines, can often be used to answer certain "what-if" questions and to make suggestions. Compared to MP alone, a decision support system usually provides greater flexibility, can deal with a far wider range of practical issues, allows for its users' local knowledge and inspiration, and attempts to enhance the powers of its users rather than replacing them by outsiders.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> Conversely, MP on its own is saddled with many limitations that make it less than ideal for solving large-scale models of practical industrial scenarios, particularly when addressing complicated problems with conflicting objectives. For example, MP sometimes provides a single, "optimal" answer; it attempts to optimize according to a single, unrealistic criterion; it is based on a model that fails to capture many of the realities of the practical situation and includes only hard (non-violatable) constraints; and it cannot provide answers within reasonable computational time (and certainly not online). Furthermore, MP requires large-scale computer hardware, software and data input, it is inflexible, and it ignores the local knowledge and inspiration of users.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> Despite this seemingly damning list of objections, MP has been, and will continue to be, a valuable part of the OR practitioner's toolkit. However, when some of the difficulties just listed arise, MP can be more effectively employed when embedded within a decision support system.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> DSS has emerged as a computer-based approach to assist decision-makers in addressing semi-structured problems by allowing them to access and use data and analytic models (Turban [9]). Such interactive systems are aimed at semi-structured problems, utilize models with internal and external databases, and emphasize flexibility, effectiveness and adaptability. These characteristics have guided much of the research in the DSS area, but the potential benefits of the DSS in the business environment have not yet been fully realized. Nevertheless, many successful DSS applications have been reported in the literature (Couillard [2]). Most of these are either large-scale systems built to facilitate well-defined and repetitive decision tasks, or else they are small PC-based systems offering quick and economic routines to support one-time decision-making (Islei et al. [7]).<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> Although the definition of the DSS concept has been elusive (Bonczek Holsapple and Whinston [1] and Er [3]), the field has flourished with the development of computer technology. Keen [8] reviewed a decade of DSS development and concluded that there is a need for a balance between each of the three DSS elements: decision, support and systems. He maintained that more research effort on the decision component was required to restore this balance, as the technology for the system component was no longer a bottleneck. To achieve "the mission of the DSS - to help people to make better decisions," Keen stressed the need for an active supporting role for "decisions that really matter."<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> Despite the proliferation of PC-based systems, the potential benefits of these systems as aids to decision-making have not been fully realized. While some DSSs may have an impact on individuals and organizations, the adoption and acceptance of these systems among decision-makers has been limited. This may be due to the inflexibility in the systems as well as their narrow design. Therefore, it is important to understand the environment of the decision-makers and the type of support they need in order to make effective decisions, and to examine the models appropriate for addressing their problems.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> Many DSSs have a basic structure that is illustrated in Figure 1. Clearly, they could include all appropriate models and their companion solution techniques that may be useful in order to gain insight into the scenario for which the particular DSS is designed. These models and techniques may not necessarily be confined to the classical OR deterministic models such as linear, integer, nonlinear and dynamic programming. They may also draw insight from such areas as queuing, scheduling, inventory and others.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> <IMG SRC="art/foulds-f1.jpg" WIDTH="400" HEIGHT="166" HSPACE="0" VSPACE="0" BORDER="0" ALT="Figure 1"><br> <img src="../cleardot.gif" width=1 height=1 vspace=2 border=0><br> <FONT face="verdana, arial, helvetica" SIZE=1><B>Figure 1. The structure of a decision support system.</B></FONT><BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> The models' bases are included in order to be used, as necessary, to solve certain sub-issues or precise questions that arise during the overall analysis of the main scenario. They can be invoked to answer "what-if" questions, to perform sensitivity analysis, and to provide precise solutions to sub-problems that can be modeled exactly. For example, within a vehicle routing DSS, a traveling salesman problem (TSP) model and various TSP solution techniques could be included. Then, if it has been established that a given vehicle will visit an identified list of clients, the TSP model and a TSP solution technique could be invoked to establish a least-distance tour.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> However, it must be stressed that any DSS should be much more than just a mere collection of models and solution techniques. Although this can be quite a valuable aid to the implementation of MP in terms of user friendliness and convenience, it should be only a small part of any DSS. The essence of the DSS is the user-system interface that allows the planners to experiment, input local knowledge and inspiration, deal with unstructured situations, be flexible, and allow for multiple objectives as well as soft (violatable to some degree) constraints.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> As an example, an educational course timetabling DSS may have, as its primary purpose, the identification of a feasible timetable without the optimization of any objective functions. The timetablers will "play" with the DSS, inputting various course-room-teacher-time slot combinations; noting various statistics that the DSS displays. Judgements as to the worth of various combinations are often made on grounds that are difficult to quantify and virtually impossible to model. Nevertheless, suggestions can be made by the DSS, based on various assignment, matching and allocation models and solution techniques from the bases in Figure 1.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=12 border=0><br> <FONT face="verdana, arial, helvetica" SIZE=3><B>Three DSS Case Studies</B></FONT><BR> <IMG SRC="../barbluehook.gif" WIDTH="468" HEIGHT="4"><br> <img src="../cleardot.gif" width=1 height=1 vspace=2 border=0><br> <I><B>Layout Manager: A Facilities Layout DSS</B></I> (Foulds [4]). Computer Automation (Ireland) ltd. (CAIL) manufacturers printed circuit boards and commercial and industrial computers at an industrial estate in Dublin. The company began experiencing difficulty in matching the productivity of its rivals in the highly competitive microcomputer manufacturing market. Part of the cause of the decline in productivity was due to inefficiencies in the mechanical assembly (MA) area, where the computer parts are actually put together to create the final products. The movement of workers, parts and equipment necessitated by the assembly sequences created a significant problem in MA. The total distance traveled by each of the entities was grossly excessive, due to poor spatial allocation of the MA operations and facilities.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> Traditionally, the most effective layout for the production schedule was found by shuffling templates. However, rapidly increasing changes in the schedule rendered this approach less and less effective. Management invited the first author to study the MA operation in order to recommend areas for improvement in its physical layout. Recognizing the importance of judgmental considerations and the need for rapid layout design decisions in a volatile environment, we felt that current layout computer programs offered little benefit. Instead, we decided to develop a DSS to address the MA layout design issue. One of the first tasks in this development was to establish appropriate criteria by which the productivity of proposed layouts could be judged. Different product mixes and levels require different criteria involving facility location and transportation times. The DSS addresses the issue of multiple objective criteria in a semi-structured decision environment.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> <I><B>FleetManager: A Vehicle Routing DSS</B></I> (Basnet, Foulds and Igbaria [6]). FleetManager is a successful milk tanker routing DSS that has generated a spectrum of benefits for the Westland Dairy Company in Hokitika, New Zealand. Much of the benefit is derived from the characteristics and attributes of a classic DSS that supports, but does not replace, the decision-maker. It does not try to provide the "answers," nor does it impose a predefined sequence of analysis. It supports semi-structured decisions where parts of the analysis can be systematized for the computer, but where the decision-maker's insight and judgement are improved.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> The DSS combines modeling techniques with database and presentation techniques. It emphasizes ease of use, user-friendliness, user control, and flexibility and adaptability. It supports all phases of decision-making. It interacts with other computer-based systems, mainly with the company mainframe system to download and upload information. The success of FleetManager can be attributed to: 1. extensive user participation and involvement, 2. an evolutionary approach to system development, 3. flexibility and simplicity of system architecture, 4. a committed and informed sponsor (the Transport Office manager), 5. accessibility and transferability of models and data, 6. availability of graphics, and 6. clarity of insights by using both judgmental and analytical models. Good relationships and close collaboration between the potential users and developers enhanced the development and implementation of the system.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> Westland's use of FleetManager has improved its vehicle-scheduling process by using existing staff and vehicles more efficiently, thereby reducing costs. In addition, it has allowed time for the schedulers and managers to improve productivity and manage people instead of running shifts and schedules manually. It provides powerful tools to create schedules, choose plans, generate alternative plans, and to assess alternative plans with respect to the given criteria. The system allows the scheduler to allot vehicle routes automatically, minimize the total distance traveled and override routes created manually. It allows for more than one source or destination, skip-a-day clients and multiple shifts, as well as empowering fine-tuning.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> The system has generated tangible and intangible benefits. In addition to the reduction in labor costs, it can benefit the schedulers by: fine-tuning the existing schedules; creating entirely new schedules; strategically planning for new sites; utilizing the fleet more efficiently; and rendering flexibility to plan for, and cope with, unexpected situations. The DSS also allows the scheduler to carry out an ad hoc analysis through "what-if" queries. It provides the scheduler with a better understanding of the business since the system alerts the users to illogical outcomes, such as unvisited suppliers and overloaded tankers. This DSS allows MP to be used to address technical sub-problems, while giving the schedulers flexibility and control concerning more broad issues that are notoriously difficult to model.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> <I><B>SlotManager: A University Timetabling DSS</B></I> (Foulds and Johnson [5]). SlotManager is a DSS that has recently been developed and implemented at the Waikato Management School to improve the course timetabling process. Although it has been developed in the context of a particular university and school, the only significant constraint on its general application is that the teaching situation being timetabled operates on a slot basis. That is, the teaching week is partitioned into a number of three-hour and four-hour slots. Each course is assigned to exactly one slot, for example, Monday, Wednesday and Friday at 9 a.m. It does not automate the process; instead, it helps timetablers by providing powerful and wide-ranging tools to designate the required slots, allocate courses to feasible slots and rooms, and create a wide variety of reports on the outcome of the process.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> Some of the benefits of the DSS can be measured, such as the time taken by the timetablers to produce an acceptable timetable. The system has clear benefits, even if they are not easily quantified. Usage of the increasingly scarce resource of teaching rooms has demonstrably improved, and there is clearly a greater level of acceptance on the part of both students and teachers of the perceived quality of the resulting timetable. Furthermore, when the inevitable disputes arise over the slots or rooms allocated to courses, these are more efficiently handled by a system that can show the alternatives (or lack of them). In particular, the "what-if" questions concerned with a perceived unsatisfactory allocation can be much more effectively explored and, if possible, resolved.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> SlotManager is designed to augment the expertise of an experienced timetabler, usually in a situation where timetables evolve from one year to the next. The reactions of experienced timetablers who have used SlotManager have been, on the whole, positive and it has been accepted as an indispensable aid to timetabling at The Waikato Management School. The DSS addresses the issue of a lack of an objective function in an unstructured decision environment.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=12 border=0><br> <FONT face="verdana, arial, helvetica" SIZE=3><B>OR and Recent Advances in Information Technology</B></FONT><BR> <IMG SRC="../barbluehook.gif" WIDTH="468" HEIGHT="4"><br> <img src="../cleardot.gif" width=1 height=1 vspace=2 border=0><br> We live in a world in which technological and social change is extraordinarily rapid. We have global public policy issues such as global warming, ozone depletion and acid rain. In the industrial arena, we have global competition spurred by technological advances and international market forces. This environment creates many opportunities for OR approaches to contribute to industrial success in the contemporary world of business.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> One of the messages that OR practitioners can convey in this context is the need for flexibility. One way to make factories more flexible is through leading-edge information technology. One approach is to harness OR to plan, schedule and control factory operations. This can lead to the implementation of processes that are so simple, standardized and reliable that they can be managed visually. In addition, information technology that links final sales points to manufacturing systems (JIT systems) is coming into its own. One challenge is to improve, and ultimately make obsolete, the heavy-handed, dysfunctional information systems that have been running many plants for decades.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> Western industry, especially the service sector, has recently experienced a widespread push toward empowering people at all levels of industrial organizations. The empowering movement has been accompanied by a pervasive focus on quality and continuous improvement. Success in global competition requires a new degree of attention to human resources.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> The reduction of lead times in new product development can provide a competitive edge. Some companies are pushing for "simultaneity," that is, parallel but coordinated activity, often in different physical locations. At the same time there is a desire for flatter organizations with fewer managerial layers, and for the replacement, at least in part, of hierarchical organizations by more market-like structures.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> Many global companies wish to take advantage of their geographical dispersion to tap knowledge and skills throughout the world. Some actually do this by implementing what is known as a trans-national strategy or structure whereby organizational learning that takes place at one location is captured and exploited at the firm's other locations.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> Where does OR fit in all this? Hopefully, everywhere, but it will not happen automatically. There is no room for complacency. However there are certain new opportunities that relate only peripherally to early perceptions of industrial management. One of these ideas involves empowering the organizational front line. In the past, OR practitioners have built systems for end-users largely as extensions of themselves. In other words, they solved the problem and delivered the solution to the potential end users.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> A different goal is to give people systems with which they can creatively solve their own problems. OR can boost industrial productivity by developing such tools for salespeople, factory workers and others. In particular, if OR can provide them with the means to monitor, understand and continuously improve their own performance, real gains can follow. These gains are now being accrued in marketing, manufacturing, the services and financial operations.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=12 border=0><br> <FONT face="verdana, arial, helvetica" SIZE=3><B>Drawing Conclusions</B></FONT><BR> <IMG SRC="../barbluehook.gif" WIDTH="468" HEIGHT="4"><br> <img src="../cleardot.gif" width=1 height=1 vspace=2 border=0><br> What conclusions can we draw for industry? Efficiencies and improved service levels provided by OR in operational systems are essential to the firms using them and cannot sensibly be abandoned. Airlines will continue to develop better systems to reallocate planes after air traffic disruptions. Foodstuff producers will build better systems to analyze bar code data on supermarket products. These systems will lead to more efficient operations, marketing, consumer products and, of paramount importance, enhanced customer value.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> We shall have better integrated manufacturing systems because of multiple levels being tied into computerized information systems. However, if OR is to remain relevant, it must take on new challenges. This gets us back to a point made earlier: empowering front-line people. Here are some examples. Some firms have huge new databases on sales and markets. Sometimes they need to be dispersed to the field and made useful to local sales people, who can combine OR and information technology with their own special knowledge of local conditions. The sales people need data analysis systems, on command, to provide fact-based dialogue with their customers.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> Another example is a system that empowers the ultimate customer. For instance, a forest products manufacturer is stimulating down-stream demand by making it easier for do-it-yourself customers to plan their own home remodeling projects. Such customers who want to build, say, an outside wooden deck attached to their house, can easily design it, using the system at a kiosk in the hardware store. The deck is visually displayed and manipulated with a mouse, with costs monitored as changes are made. After the design is complete, a push of a button prints the final result and a bill of materials. Another opportunity lies in supporting the goal of simultaneity and time compression. In this regard, it is important to provide tools to assist cooperative work. If we are to shorten lead times, multiple persons will have to work on different aspects of the same problem, at the same time, but in a coordinated way - computer-assisted cooperative work supported by group decision support systems.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> The complexity of industrial enterprise now requires end-user focussed information systems, often computerized. These are necessary in order for organizations to function efficiently and effectively. The challenges include: the assimilation of information systems so that local knowledge can be used easily, support for flat organizations and cooperative work, and the empowerment of front-line people and making them semi-independent. These developments are significant for the evolution of OR. If they are embraced effectively, then OR is likely to have even more impact on economic development and the quality of life.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> <CENTER> <TABLE BORDER=0 CELLSPACING=0 CELLPADDING=6 width=400> <TR> <TD BGCOLOR="#660000"><FONT face="verdana, arial, helvetica" SIZE=3 color=#FFFFFF><CENTER><B>Historical Perspective</B></CENTER></TD> </TR> <TR> <TD BGCOLOR=#CCCC99><FONT face="verdana, arial, helvetica" SIZE=2> Operations research had its beginnings in World War II, when British military authorities faced pressing problems that were not amenable to any known scientific methods. These problems arose from the development of new technology and its application to military operations, such as the location of radar installations and the detection of submarines. To make the best use of the hardware and to analyze the data that it produced, British officials created teams of scientists from a wide variety of backgrounds. <BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> After the war, British and American scientists realized that this approach of creating diverse teams to build and optimize models could be used to great effect in the world of commerce and public administration. The practice spread throughout the developed world. With spectacular increases in computer power, more complex models have been progressively analyzed over the decades. <BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> The modeling/optimization approach was a large part of OR theory and practice throughout the 1960s and '70s. Starting in the early 1980s, however, many OR leaders concluded that this approach was not adequate in all instances. Criticism came from, among others, Churchman, Machol and Ackoff in the United States, Daellenbach in New Zealand, and Beer, Eden and Checkland in the United Kingdom. Many of these people suggested systems thinking, soft systems methodology and related techniques were more appropriate than mathematical programming (MP) for many complex, ill-defined, difficult-to model issues that OR practitioners were beginning to face. <BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> The 1990s saw a revolution in computer and information technology that has had a significant effect on OR practice. Today, there are less specialized OR groups within organizations and very few groups with the diverse range of skills that were seen in the 1940s. A much wider range of individuals are now considered OR practitioners, even if they themselves do not realize it. In the last decade, we have also seen the rise of the spreadsheet, easy access to inexpensive OR software, simple communication and the spread of OR knowledge via the Internet and other forms of modern information technology. This led to more individuals than ever before being engaged in OR, often alone, and sometimes with little OR training or education.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> <B> Les Foulds and Cherian Thachenkary</B> <BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> </font> </TD> </TR> </TABLE></CENTER> <img src="../cleardot.gif" width=1 height=1 vspace=12 border=0><br> <FONT face="verdana, arial, helvetica" SIZE=3><B>References</B></FONT><BR> <IMG SRC="../barbluehook.gif" WIDTH="468" HEIGHT="4"><br> <ol><FONT Size=2> <LI> R.H. Bonczek, C.W. Holsapple and A.B. Whinston, "Foundations of Decision Support Systems," Academic Press, New York, 1981. <LI> J. Couillard, "A decision support system for vehicle fleet planning," <I>Decision Support Systems,</I> Vol. 9 (1993), p. 149-159. <LI> M.C. Er, "Decision support systems: A summary, problems, and future trends," <I>Decision Support Systems,</I> Vol. 4 (1988), p.355-363. <LI> L.R. Foulds, "LayoutManager: A microcomputer-based DSS for facilities layout," <I>Decision Support Systems,</I> Vol. 20 (1997), p. 199-213. <LI> L.R. Foulds, D.G. Johnson, "SlotManager: A Microcomputer-based Decision Support System for University Course Timetabling," <I>Decision Support Systems,</I> Vol. 27 (2000), p. 367-381. <LI> C.B.Basnet, L.R. Foulds, and M.Igbaria, "FleetManager: A microcomputer-based decision support system for vehicle routing," <I>Decision Support Systems,</I> Vol.16 (1996), p. 195-207. <LI> G. Islei, G. Lockett, B. Cox, S. Gisbourne, and M. Stratford, "Modelling strategic decision-making and performance measurement at ICI Pharmaceuticals," <I>Interfaces,</I> Vol. 21, No. 6 (1991), p. 4-22. <LI> P.G.W. Keen, "Decision support systems: The next decade," <I>Decision Support Systems,</I> Vol. 3 (1988), p. 253-265. <LI> E. Turban, "Decision Support and Expert Systems: Management Support Systems," 3rd edition, Macmillan, New York, 1993. </FONT> </OL> <IMG SRC="../barbluehook.gif" WIDTH="468" HEIGHT="4"><br> <img src="../cleardot.gif" width=1 height=1 vspace=2 border=0><br> <FONT face="verdana, arial, helvetica" SIZE=2><I>Les R. Foulds (<A HREF="mailto:lfoulds@waikato.ac.nz">lfoulds@waikato.ac.nz</A>) is a professor in the Department of Management Systems, University of Waikato, New Zealand. Cherian S Thachenkary (<A HREF="mailto:thachenkary@gsu.edu">thachenkary@gsu.edu</A>) is an associate professor in the Department of Management, The J. Mack Robinson College of Business, Georgia State University, in Atlanta. </i><BR>  <img src="../cleardot.gif" width=1 height=1 vspace=5 border=0><br>  <img src="../barblue.gif" width=468 height=1 border=0 alt=" "><br> <img src="../cleardot.gif" width=1 height=1 vspace=8 border=0><br> <font face="verdana, arial, helvetica" size="1"> <CENTER><script type="text/javascript"></script> <script type="text/javascript" src="http://pagead2.googlesyndication.com/pagead/show_ads.js"> </script></CENTER><br> <img src="../cleardot.gif" width=1 height=1 vspace=8 border=0><br> <FONT face="verdana, arial, helvetica" SIZE=1> <LI> <A HREF="tocfr.html" Target="_top"><I>Table of Contents</I></A><br> <img src="../cleardot.gif" width=1 height=1 vspace=4 border=0><br> <LI> <A HREF="../../ORMS.shtml" TARGET="_top"><I>OR/MS Today</I> Home Page</A><br>  <img src="../cleardot.gif" width=1 height=1 vspace=8 border=0><br> <img src="../barblue.gif" width=468 height=1 border=0 alt=" "><br> <b><i>OR/MS Today</i></b> copyright � 2001 by <A HREF="http://www.informs.org/" target="_new">the Institute for Operations Research and the Management Sciences</a>. All rights reserved.<br> <img src="../cleardot.gif" width=1 height=1 vspace=3 border=0><br> <img src="../bargrey.gif" width=468 height=1 border=0 alt=" "><br> <b>Lionheart Publishing, Inc.</b><br> 506 Roswell Street, Suite 220, Marietta, GA 30060, USA<br> Phone: 770-431-0867 | Fax: 770-432-6969<br> E-mail: <a href="mailto:lpi@lionhrtpub.com">lpi@lionhrtpub.com</a><br> URL: <A HREF="http://www.lionhrtpub.com" target="_top">http://www.lionhrtpub.com</A><br> <img src="../cleardot.gif" width=1 height=1 vspace=3 border=0><br> <img src="../bargrey.gif" width=468 height=1 border=0 alt=" "><br> Web Site � Copyright 2001 by Lionheart Publishing, Inc. All rights reserved. </td></tr></table> <img src="../cleardot.gif" width=1 height=1 vspace=2 border=0><br> </BODY>