Algorithms for an Autonomous Car

Professor Edwin Olson

University of Michigan

Department of Electrical Engineering and

Computer Science

Computer and Science Engineering

Abstract:

This talk will describe MIT's DARPA Urban Challenge entry "Talos", which finished the race with the fourth fastest time. DARPA asked entrants to build autonomous cars capable of driving in urban environments amidst other traffic, presenting many challenges to perception, planning, and control. This talk will describe these systems, highlighting some of the clever ideas and unique choices made by our team. Interesting open problems remain, as highlighted by some of the shortcomings of our system.

Speaker’s Bio:

Edwin Olson is an Assistant Professor in the EECS department at the University of Michigan. His research focuses on robust robotic sensing, environment modeling and mapping, and motion planning. Increasing human acceptance of robots is also a strong interest: he is currently incorporating machine learning into a motion planner so that robots can behave in a more natural, human-like way.

Friday, November 14, 2008

3:30 – 4:30 p.m.

Rm. 1500 EECS