Hugo Simao and Warren Powell won the 2010 Best Paper Prize from the Society for Transportation Science and Logistics at the Informs annual meeting this fall. The paper, entitled "An Approximate Dynamic Programming Algorithm for Large-Scale Fleet Management: A Case Application" uses approximate dynamic programming to solve a problem with an infinite-dimensional state space, and high-dimensional action and outcome spaces. If it were formulated as a deterministic optimization problem, it would produce an integer program with over 10^23 integer variables and constraints. The model was able to accurately capture the collective intelligence of the sophisticated dispatch operation at Schneider National, one of the largest truckload motor carriers in the U.S.
Nov. 15, 2010