Focus on new faculty: JIM LUEDTKE
 ssistant Jim Luedtke never thought he’d have the chance to return to UW-Madison. The Colby, Wisconsin, native and ISyE undergraduate alumnus joined the UW-Madison faculty in fall 2008 after obtaining his PhD from Georgia Institute of Technology and completing a postdoc at IBM Research.
While the decision to settle in Wisconsin was easy, since he has family in the state and has always wanted to return, not all of the situations Luedtke encounters in his optimization research are so clear-cut. He is investigating two challenging optimization situations: making discrete decisions and making decisions in the face of uncertain data and outcomes.
Luedtke is collaborating with Associate Professor Jeff Linderoth on an integer programming project to tackle discrete decisions,
which are essentially yes or no situations. An example of a discrete decision is when a retail manager decides whether to open a new warehouse. “It’s an all or nothing decision—you can’t open a third of a warehouse,” Luedtke says. He and Linderoth are studying
algorithms to solve problems with discrete decisions that are further complicated by nonlinear constraints.
Luedtke is also working on stochastic optimization, which addresses uncertainty in problem solving. In traditional stochastic optimization research, the objective is to find the best solution on average. Often, this assumes a decision will be made multiple times. However, Luedtke recognizes that in many situations, a decision is made only once, and the decision-maker cares about a single outcome rather than an average of multiple outcomes. An example is investing in stocks: When an investor retires, they care about their personal outcome rather than the average outcome of all investors.
“Decision-makers want to avoid bad outcomes, and I’m researching methods that try to find good solutions that limit the likelihood or magnitude of bad outcomes,” says Luedtke. “It’s about finding a better match between the decision and the risk tolerance of the decision-maker.”
He is also exploring how his techniques can be applied. One example is Luedtke is part of a new partnership between the UW-Madison School of Human Ecology Center for Retailing Excellence, with which Luedtke is affiliated, and Hy-Vee Inc., an employee-owned retailer and grocer that will be opening two stores in the Madison area.
The Hy-Vee project requires Luedtke to examine uncertainty around product demand in order to create a new warehouse process. The new process will be based on the flow through philosophy that moves product from the supplier to the warehouse to the store quickly. Luedtke will use stochastic and discrete optimization in a system that can operate the new process.
| |
