ORMS Today - Life After Edelman - Public Sector

<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> - December 2003</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 SIZE="+2"><B>Life After Edelman</B></FONT><BR> <img src="../cleardot.gif" width=1 height=1 vspace=2 border=0><br> <FONT SIZE="+1"><I>They once played in the "Super Bowl of OR," but where are they now? Researchers track down nearly two dozen former Edelman finalists in the not-for-profit sector to see if their contributions stand the test of time.</I></FONT><BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> <FONT face="verdana, arial, helvetica" SIZE=2><B>By Peter C. Bell</B></FONT><br> <hr size="1"> <img src="../cleardot.gif" width=1 height=1 vspace=12 border=0><br>    The Franz Edelman prize competition is a showcase for the very best applications of operations research/management science. We set out to find out what had happened to the private sector finalists from the 1989-1998 competitions following presentation and publication of these very effective examples of OR/MS practice. We have published the results of this work in articles in <I>Operations Research</I> [Vol. 51, No. 1, January-February 2003], <I>Interfaces</I> [Vol. 32, No. 2, March-April 2002] and <I>OR/MS Today</I> [Vol. 28, No. 3, June 2001]. <P>Our research often prompts the question: What about the not-for-profit applications? This sector is a very important client for OR/MS, with about one-third of the Edelman finalist applications in the not-for-profit category. We ignored not-for-profit applications in our earlier work because our research objective was to find examples of OR/MS work that met a definition of "strategic OR/MS" as "OR/MS work that creates a sustainable competitive advantage" for the firm [Bell, <I>JORS,</I> 1998]. We found that about half (20 of 42) of the private sector applications met this definition of strategic OR/MS. Significantly, 18 of the 20 applications identified as "strategic" were still in regular use more than five years after initial implementation. The two that were no longer in use had had a useful life of more than 15 years and were only discontinued when a change of corporate strategy rendered them obsolete. <P>Sustaining a competitive advantage is usually not a critical objective in the not-for-profit sector (although it may be applicable in the military!), and we did not expect to find support for strategic OR/MS in this sector. However, the obvious interest in the fate of the non-corporate Edelman finalist applications prompted us to further examine these works. We followed the same methodology as earlier: We first studied the written documentation for the 1989-1997 finalists, and then attempted to conduct telephone interviews with at least two people familiar with the history of each application since the Edelman competition. Unlike the private sector, where we had 100 percent success with our contacts, we do not have recent information on all the applications (although we are still trying).<BR> <img src="../cleardot.gif" width=1 height=1 vspace=12 border=0><br> <FONT face="verdana, arial, helvetica" SIZE=3><B>The not-for-profit applications</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> Twenty-three Edelman finalist articles published in <I>Interfaces</I> between 1990 and 1998 featured a client organization from the not-for-profit sector. Three of these are military applications (Israeli Air Force and two from the U.S. armed forces), leaving a highly diverse group of 20 non-military, not-for-profit applications. The most common pure "public sector" applications are projects done for government by a team made up of people from government and academia. Examples include the New Haven needle exchange (1993), NASA (1994) and Hanshin Expressway (1995). The remaining applications all have private sector involvement  many of the projects were done for the public sector by private sector groups. Example include Sabre's yield management for SNCF, GIRO's HASTUS system for crewing public transit systems and Deloitte & Touche Consulting's planning model for the South African National Defense Force. <P>Close examination of the private sector Edelman finalist applications, including conversations with the people who built and have worked with these very successful OR/MS applications, is a "feel-good" experience for anyone interested in OR/MS. In the public sector, we found the "feel-good" experience to be, if anything, even stronger; many of these works have had a major impact. Here are five examples that illustrate different kinds of sustained impact from some public sector Edelman finalist applications. <P>1. Bus crew scheduling software, HASTUS [<I>Interfaces,</I> 1990], began life as a student project under the direction of Professor Jean-Marc Rousseau at the University of Montreal. The first customer was STCUM (The Quebec City Transit Company) which initially saved about $3 million annually in manpower costs and $1 million annually in vehicle costs. STCUM now has about 10 schedulers using HASTUS daily, and its savings have increased to around $6 million annually. Rousseau formed a company, Giro, to further develop and market HASTUS, and the software is now in use in more than 200 cities around the world including New York and Singapore. New York has between 3,000 and 4,000 buses, but each line is operated separately so the 4,000-bus problem is decomposed into some 100 smaller problems. Probably the largest single problem is Ottawa where 1,000 buses are crewed simultaneously. <P>Says Rousseau: "I think a company that has been using any software for doing their schedule for the past five or six years is probably completely unable to revert back to manual. In fact, the schedule is the product of a transit company. It is what they are selling, their schedule. It is really at the base of the operation; 70 percent or 80 percent of their cost is related to the schedule." <P>HASTUS was completely redesigned in the 1990s, but the basic solution procedure for large problems remains the heuristic method presented in the Edelman competition. Smaller cities can now make use of a new optimization algorithm based on column generation and set covering. Giro now employs about 140 people, including 40 who develop the software on a continuous basis. Competition has appeared from groups in Australia, Spain, Germany and France, but these tend to remain small and in a survival mode. "Every time we do a test we produce better solutions than our competitors," says Rousseau, who believes Giro has the best method on the market. <P>2. Edward Kaplan and Elaine O'Keefe's study of the New Haven needle exchange [<I>Interfaces,</I> 1993] has been reviewed multiple times since the Edelman competition by groups as varied as the General Accounting Office (GAO), the Center for Disease Control and the National Academy of Sciences, and by individual scholars. The assumptions have been relaxed or complicated, but the original results have been reinforced. A recent GAO review concluded that the benefits of needle exchange were even greater than Kaplan and O'Keefe had suggested. Kaplan and O'Keefe's modeling methods have precipitated a number of new studies in the United States and Eastern Europe that typically made estimates of the numbers of infections averted. The results have been very consistent. <P>The New Haven needle exchange program has not been studied for the last several years, but important changes have occurred. For example, it is now possible to purchase syringes without a prescription, and many addicts now do this rather than use the needle exchange. <P>There are now well over 100 needle exchange programs in North America, and the results developed by Kaplan and O'Keefe continue to inform the debate; their work was recently cited in Saskatoon, Canada, to try and persuade the city council that a needle exchange was a good idea. <P>3. The Aiden system [<I>Interfaces,</I> 1992] continues to be used seven days a week by the U.S. Military Airlift Command for aircraft scheduling. Aiden was rushed into service some nine months ahead of schedule in response to the need to transport the forces to be used for Desert Storm. After implementation and following successful use for the Desert Storm airlift, the system was further developed to include several new components. Additions include a special assignment airlift mission's component which allows the Army to move troops following redeployment, and an air refueling part of Aiden that schedules the tanker airplanes that refuel fighters, bombers and the airlifters. In 1996, the Aiden software was turned over to a contractor for maintenance while the developers were retained in an advisory/consultant role. <P>The Aiden system supported reorganization and centralization of the command by enabling a few planners at a central location to handle scheduling that had been done partially at a number of air bases along both coasts, in Germany and in Hawaii. Aiden is now just a part of the way the air mobility command fulfills its mission. The command not only uses it to schedule its day-to-day operations, but it also employs Aiden to develop war plans and help planners figure out how much fuel they need to have positioned at various bases around the world in order to defend a particular country or theatre. Aiden is used to schedule aircraft in "paperwork exercises" where Aiden does the scheduling as planners simulate the airplanes flying. Says one official: "It is just part of the way the command does its business There is no option about using it." <P>The success of Aiden has provided the development group with the opportunity to undertake several new projects, including work for the Army Corps of Engineers and other military and nonmilitary sponsors. <P>4. The Hanshin Expressway in Osaka, Japan [<I>Interfaces,</I> 1995] initiated an automatic traffic control system in 1970 that uses linear programming to calculate the maximum allowable inflow at each entry ramp every five minutes. The system has operated effectively for more than 30 years and has been developed from a system using one computer to the present 19-computer system. A new traffic control system is under construction and includes a simulation model named HEROINE (Hanshin Expressway Real-time Observation-based & Integrated Network Evaluator) for online traffic forecasting. A recent survey found that almost 100 percent of drivers using the expressway use the real-time displays to plan their routes. <P>5. No discussion of the not-for-profit Edelman finalist application would be complete without a mention of diagnosis related groups (DRGs) [<I>Interfaces,</I> 1990]. DRGs, developed by Robert Fetter (OR Department at Yale University) to provide product definitions for the output of hospitals, were adopted by Medicare in 1983. By 1990, DRGs had saved the U.S. government more than $50 billion in Medicare hospital payments. Today, the U.S. medical payments systems would collapse without DRGs; almost everyone in health care management in the United States is thoroughly familiar with DRGs. In addition, more than 30 other countries use DRGs as the basis for medical payments, hospital budgets and quality control.<BR> <img src="../cleardot.gif" width=1 height=1 vspace=12 border=0><br> <FONT face="verdana, arial, helvetica" SIZE=3><B>Classifying not-for-profit applications</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> In our private sector research, we grouped OR/MS applications into four categories: <UL> <LI> "strategic applications" that create a sustainable competitive advantage for their host firms; <LI> "insufficient impact" applications where an advantage is gained but given the size of the firm, the dollar value is too small to be considered strategic; <LI> "contributor" applications where the OR/MS work helps the firm make one or more important (perhaps "strategic") decisions, but the major benefit arises from the impact of the strategic decision(s) that was made; and <LI> "non-sustainers" where the firm gains a benefit, but the OR/MS work appears to be easily replicated by competitive firms and so the advantage gained is short lived. </UL> <P>Striving for a competitive advantage is not an objective of the not-for-profit sector, nor does cost cutting have the same importance as in the private sector. We do, however, find not-for-profit analogues of three of our four private sector categories of application, and these are summarized in Table 1. The fourth category, the "non-sustainer," appears to have no general not-for-profit sector equivalent although there may be analogous military applications. "Non-sustainer" applications are open to easy replication by competitors, but most of the not-for-profit sector is not driven by competition. Such applications may exist in the military (for example, war game simulations, where one force developed computer-based battlefield simulations, but as soon as the idea became public, every military developed generally similar software).<BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> <TABLE BORDER="1" CELLSPACING="0" CELLPADDING="6"> <TR ALIGN="center" VALIGN="Top"> <TD><FONT face="verdana, arial, helvetica" SIZE=2><B>Private sector categories of OR/MS applications</B></TD> <TD><FONT face="verdana, arial, helvetica" SIZE=2><B>Not-for-profit equivalent categories of OR/MS applications</B></TD> </TR> <TR ALIGN="Left" VALIGN="Top"> <TD><FONT face="verdana, arial, helvetica" SIZE=1><DL> <DT>Strategic OR/MS applications <DD> Created a sustained advantage. </DL> </TD> <TD><FONT face="verdana, arial, helvetica" SIZE=1><DL> <DT>"Stayers" <DD> OR/MS applications that were used regularly over a substantial period of time to help manage a critical activity. </DL> </TD> </TR> <TR ALIGN="Left" VALIGN="Top"> <TD><FONT face="verdana, arial, helvetica" SIZE=1><DL> <DT>Contributors <DD> OR/MS work used to make an important decision but the decision produced the payoff, not the OR/MS. </DL> </TD> <TD><FONT face="verdana, arial, helvetica" SIZE=1><DL> <DT>"Policy formulators" <DD> OR/MS studies that formulated lasting public policy. </DL> </TD> </TR> <TR ALIGN="Left" VALIGN="Top"> <TD><FONT face="verdana, arial, helvetica" SIZE=1><DL> <DT>Insufficient impact <DD> The benefits of the work were not large enough to be "strategic." </DL> </TD> <TD><FONT face="verdana, arial, helvetica" SIZE=1><DL> <DT>"Low impact" <DD> The benefits of the work were not large enough to be memorable. </DL> </TD> </TR> <TR ALIGN="Left" VALIGN="Top"> <TD><FONT face="verdana, arial, helvetica" SIZE=1><DL> <DT>Non-sustainers <DD> The OR/MS work looks easily replicated and any advantage eroded. </DL> </TD> <TD><FONT face="verdana, arial, helvetica" SIZE=1><DL> <DT>"Non-sustainers" <DD> May be limited to the military where "competition" exists in the not-for-profit sector </DL> </TD> </TR> </TABLE> <img src="../cleardot.gif" width=1 height=1 vspace=2 border=0><br> <FONT face="verdana, arial, helvetica" SIZE=1><B>Table 1: Private sector categories of OR/MS applications and the not-for-profit sector equivalent categories.</B></font><BR> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> <P>Our examination of the not-for-profit Edelman finalist applications from the 1989-1997 prize competitions suggests these applications should be classified as "stayers": <UL> <LI> <B>HASTUS</B> [<I>Interfaces,</I> 1990]. HASTUS has been in regular use since 1990 at a host of sites to perform a critical function for public transit companies. <LI> <B>Hidroelectrica Espanola</B> [<I>Interfaces,</I> 1990]. The models developed to manage a system of reservoirs have been improved and developed and are still in regular use. <LI> <B>US MAC</B> (Aiden) [<I>Interfaces,</I> 1992]. Aiden has been used seven days a week to schedule aircraft for more than 10 years. <LI> <B>USPS</B> [<I>Interfaces,</I> 1992]. The original model was in use up to a few years ago with new versions now in development. <LI> <B>Hanshin Expressway</B> [<I>Interfaces,</I> 1995]. Optimization has been continuously used to schedule this critical Osaka expressway for more then 30 years. <LI> <B>PONTIS</B> [<I>Interfaces,</I> 1997]. Now a national system used in 48 states, run several times a year and consulted almost daily to plan operations. <LI> <B>UK National Health Service</B> [<I>Interfaces,</I> 1997]. The model is used annually to allocate some �25 billion to health organizations in England, and has been regularly maintained and updated. <LI> <B>SNCF</B> [<I>Interfaces,</I> 1998]. The scheduling and yield management models developed for the SNCF TGV trains continue to be a key part of their operation. </UL> <P>These finalists meet our definition of "policy formulators": <UL> <LI> <B>DRGs</B> [<I>Interfaces,</I> 1990]. A single OR/MS study identified DRGs (although many refinements have been undertaken), and DRGs are now a mainstay of medical administration and have saved the U.S. government many billions of dollars. <LI> <B>New Haven Needle Exchange</B> [<I>Interfaces,</I> 1993]. Kaplan and O'Keefe's work showed the advantage of needle exchange programs, and the resulting programs have produced huge social benefits. <LI> <B>New York City</B> [<I>Interfaces,</I> 1993] The model was built to help New York comply with a court-ordered 24 hours "arrest to arraignment." Once in compliance, the need for the model went away. <LI> <B>SANDF</B> [<I>Interfaces,</I> 1997]. Optimization was used to define an army, but the benefits came from the army. <LI> <B>Kuwait al Manakh Stock Market</B> [<I>Interfaces,</I> 1997]. Optimization backed by law resolved the $94 billion in uncovered cross-payments after the collapse of the stock market. Any continuing benefit was the result of the resolution of the issue, not the OR/MS. <LI> <B>Tritium Supply</B> [<I>Interfaces,</I> 1998]. The decision analysis model was used to provide advice to a group responsible for an important strategic decision. </UL> <P>Finally, a couple of applications do not appear to have had sufficient impact to attract the attention of senior management and have not survived. We classify them as "low impact": <UL> <LI> <B>China Coal planning</B> [<I>Interfaces,</I> 1995]. The model received World Bank funding, but when the funding ran out the modelers left and the substantial benefits that the model provided were not sufficient to maintain interest in continuing the work. <LI> <B>Reshaping U.S. Army</B> [<I>Interfaces,</I> 1995]. The work appears to have ended when one of the principals retired. </UL> <img src="../cleardot.gif" width=1 height=1 vspace=6 border=0><br> <FONT face="verdana, arial, helvetica" SIZE=3><B>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> Examination of the history of the not-for-profit Edelman finalist applications from the 1989-1997 competitions following publication of the works confirms the view that these were important OR/MS applications. The great majority of those applications where we have been able to collect longitudinal data have had a major impact, either through continual use on a regular basis to address an important not-for-profit problem, or through informing debate, often over a considerable period of time, on an important public policy issue. <BR> <img src="../cleardot.gif" width=1 height=1 vspace=4 border=0><br> <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><B>Peter C. Bell</B> (<A HREF="mailto:Pbell@Ivey.ca">Pbell@Ivey.ca</A>) is a professor in the Richard Ivey School of Business at the University of Western Ontario.</i></font><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> <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</A><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 � 2003 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 Rd., 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 2003 by Lionheart Publishing, Inc. All rights reserved. </td></tr></table> <img src="../cleardot.gif" width=1 height=1 vspace=2 border=0><br> </BODY>