Title: Robust Resource Allocation under Uncertainty for Transportation Systems
Speaker: Lavanya Marla, Carnegie Mellon University
Time: 11 am, Monday, September 10, 2012
Venue: Room 217, Mechanical Engineering
Abstract:
Abstract:
Transportation scheduling and routing problems, with their inherent complexity and large-scale nature, are ideal candidates for the application of optimization techniques. Conventional models, however, include simplifying assumptions, such as deterministic model inputs – guaranteeing that such solutions are vulnerable to inherent real-world uncertainty. In this talk, I address the fundamental question of building robust solutions, that is, solutions which (i) are less fragile to disruption, (ii) are easier to repair if needed, (iii) optimize the realized, rather than planned, problem objective.
Approaches towards robustness adopt a proactive stance towards future uncertainty, and anticipate potential disruptions in the planning stage. Iwill discuss generally-applicable modeling paradigms of robust planning, their applicability to mobility networks; and present new extended models which improve solution quality and tractability. Drawing on experience from multiple applications, I will discuss insights about the relationships between the robust models with data and robustness metrics, and solution differences between these paradigms. I also present algorithms which use, at their core, our extended models, and enable robust solution generation for large-scale network-based resource allocation at the planning stage as well as in a dynamic, real-time, setting.
Speaker Bio:
Lavanya Marla is a Systems Scientist with the iLab’s Mobility Analytics group in the Heinz College at Carnegie Mellon University; and holds a PhD in Transportation Systems from the Massachusetts Institute of Technology. She develops models and algorithms for large-scale transportation and logistics systems, using tools from optimization, statistics, simulation, simulation-optimization and other computational techniques. Her research interests include robust resource allocation for large-scale systems, decision-making under uncertainty, real-time dynamic re-planning, and multi-agent systems; with applications in aviation, logistics, emergency management and shared transportation systems. She is a recipient of the Alfred P. Sloan Foundation fellowship, the Ann Hershfang scholarship awarded by the Women’s Transportation Seminar, and is a Co-PI on a US Department of Transportation award for a University Transportation Center at CMU (Jan 2012). She will be joining the tenure-track faculty in the Department of Industrial and Enterprise Systems Engineering at the University of Illinois at Urbana-Champaign starting in Jan 2013.
News Category