From Good Medical Decisions to Better Teamwork, Pitt Industrial Engineers Study Efficiency In and Out of the Factory to Win National Awards From the Institute of Industrial Engineers

PITTSBURGH-Several researchers from the University of Pittsburgh's School of Engineering will be recognized by the world's largest organization focused on improving industrial quality and productivity. Members of Pitt's Department of Industrial Engineering will receive national awards from the Institute of Industrial Engineers (IIE), a professional society for industrial engineers, for research that includes determining the best time to make medical decisions and for reorganizing the nation's transplant organ system to save more lives. The IIE will present the awards at its annual conference in Nashville May 20.

The IIE will honor Andrew J. Schaefer with the Dr. Hamed K. Eldin Outstanding Young IE Award in Education. The award recognizes individuals in academia younger than 35 who have demonstrated outstanding characteristics in leadership, professionalism, and potential in industrial engineering. An associate professor of industrial engineering and a Wellington C. Carl Faculty Fellow, Schaefer studies the optimum time to make certain medical decisions, particularly regarding liver transplants and initiating treatment for HIV.

Assistant Professor Lisa M. Maillart won the IIE award for the best paper on quality and reliability engineering to appear in IIE Transactions, the institute's peer-reviewed journal. Published last year, Maillart's paper, "Maintenance Policies for Systems With Condition Monitoring and Obvious Failures," took a unique approach to analyzing industrial systems that break down over time.

Pitt alumnus Nan Kong won the first-place Pritsker Doctoral Dissertation Award for developing concepts that could reorganize the nation's transplantable organ procurement and allocation system so that 100 more people a year could get the transplants they need. Kong earned his PhD degree in industrial engineering at Pitt last year and recently accepted a faculty position at Purdue University.

A shared award will go to Mary E. Besterfield-Sacre, an associate professor of industrial engineering; Renee M. Clark, a visiting research assistant professor; Larry J. Shuman, an industrial engineering professor and associate dean for academic affairs in Pitt's engineering school; emeritus professor Harvey Wolfe; and Pitt graduate student Tuba Pinar Yildirim. The researchers will receive the Best Paper Award in Engineering Education at the national conference's Research Sessions, which showcase the latest research. The team investigated whether a formula largely used to evaluate industrial systems can also be used to streamline group projects in the work place.

The IIE awards illustrate the quality and growing prominence of work coming out of Pitt's industrial engineering department, said Bopaya Bidanda, the department chair and Ernest E. Roth Professor of Industrial Engineering. "These awards are indicative of the rise of the industrial engineering department at Pitt," he said, adding that the recognition also honors the Pittsburgh region, where industrial engineering had its roots in such companies as Westinghouse Electric Corporation and H.B. Maynard Co.

Founded in 1948, IIE is the world's largest society dedicated to the support of the industrial engineering profession and all people involved in improving quality and productivity. The institute has more than 15,000 members worldwide and more than 200 chapters. With headquarters in Norcross, Ga., the institute offers training courses at the facility as well as at corporate client sites and online. For information on the IIE's annual conference, visit the conference Web site at www.iieannual.org.

*******

More information on each award recipient's research is included below.

Andrew J. Schaefer works on projects that are among only a few in which mathematical probabilistic models are applied to medical decisions. In terms of transplants, Schaefer currently works with Mark S. Roberts, an associate professor of medicine in Pitt's School of Medicine, and other researchers, including Maillart, on a National Institutes of Health-funded project regarding live-donor liver transplants. One goal of the project is to determine the best time to conduct a transplant given the condition of the living donor and the recipient. (Living-donor transplants do not apply to the usual wait-list rules, so the only decision is whether to perform a transplant.) Because living-donor transplants put the donor's life at risk, a second aim of the project is to pinpoint the optimal time that a person in need of a liver could hold out for an organ from a deceased person before accepting one from a living-donor.

Schaefer applies the same model to help determine the best time for people with HIV to begin treatment, a question considered by some to be the most important in HIV therapy. Schaefer's work balances the benefits of seeking treatment in the later stages of infection-such as delaying negative side effects and preserving resources needed later-against the disadvantages, such as the risk of developing advanced AIDS and associated health problems. By considering these possibilities in the context of a person's life expectancy, current health condition, and the natural progression of HIV, all at the time a decision is being made, a person and their physician can determine whether treatment can wait or should begin. Schaefer has noted in presentations on this topic that the probability results suggest starting HIV therapy immediately, which disagrees with the current guidelines from the federal Centers for Disease Control and Prevention.

Lisa M. Maillart researched the inherent shortcomings of monitoring certain wear-related mechanical signals on a scheduled basis, a route not often taken in maintenance research. Moreover, Maillart focused on these shortcomings in settings where failures-known as "obvious failures"-are apparent to the machine operator. Other work concentrates on glitches that must be rooted out, or "silent failures." Maillart's work offered insights into the structure of the optimal way to schedule these observations when the information that they provide is noise-corrupted, or imperfect. Such results can increase the speed at which a maintenance manager can solve for the optimal schedule as well as ease this solution's implementation. Maillart recently had another paper accepted for publication on the same topic that addresses silent failures.

Nan Kong-who studied under Schaefer-developed concepts that could reorganize the nation's transplantable organ procurement and allocation system, increasing the number of transplants and general access to quality organs. As Kong detailed in his dissertation, the United States consists of nearly 60 procurement agencies broken into 11 regions. Agencies supply organs for transplant patients in their respective region. Unfortunately, the regions largely follow geographic borders, Kong said. The ratio of donors to recipients varies from region to region and some regions have access to better-quality organs. Kong applied an engineering analysis approach to compute the most efficient way to reconfigure the regions. The result suggested reorganizing the 11 regions into five to seven regions based on the highest possible number of successful transplants. The reorganization would allow for about 100 more transplants a year and equalize the regions' donor-to-recipient ratio and access to quality organs, Kong said.

Graduate student Tuba Pinar Yildirim, faculty members Mary E. Besterfield-Sacre, Larry J. Shuman, and Harvey Wolfe, and visiting instructor Renee M. Clark investigated whether a formula for evaluating industrial systems can also breakdown team projects into distinct stages. They found that these stages highlight the strengths and pitfalls of group work so that the overall task can be streamlined.

The authors began with the documented premise that American product design-typically the task of engineers-is inferior to that of many foreign companies. The authors wanted to determine if the team-based design process could be dissected-and its shortcomings exposed-with the Markov Chain, a system for observing industrial systems. In the Markov Chain, the evolution of a system's condition is broken into stages. Each stage develops in relation only to the stage immediately before it. For example, if a process began well but failed at the end, its final state is as a failure.

While observing a group of undergraduate engineering design students, the researchers took note of certain actions and the usual outcome. For instance, when one person led the group in discussion or activity, they would afterward lapse into a nonproductive state of waiting. Conversely, group discussion often spurred new ideas and better evaluation. By singling out these stages and their respective effect, the researchers concluded that the Markov Chain is an apt tool for reviewing group design projects.

###

05/16/07/scl