"Uncertainty in Mobile Robotics"

Sebastian Thrun, Carnegie Mellon University

This talk will present a family of probabilistic algorithm for mobile robotics. These algorithms have in common that they explicitly represent uncertainty using probabilistic metaphors, and that they utilize efficient statistical methods (such as EM and value iteration) to tackle high-dimensional state estimation (=learning) and planning problems. The talk will discuss application of these algorithms to mobile robot localization in dynamic environments, mapping, navigation, and active exploration. The algorithms presented in this talk were key components of a series of successful mobile robots, which were installed as interactive tour-guides in museums. The latest tour-guide, called Minerva, interacted with more than 50,000 people in a Smithsonian museum in Washington, DC. If time permits, the speaker will also discuss ongoing work on the design of a new programming language, which seeks to leverage these results to a much broader class of embedded systems.

Speaker's Bio: Sebastian Thrun is an Assistant Professor of Computer Science and Robotics at Carnegie Mellon University, with research interests in mobile robotics and artificial intelligence. Further information can be found at http://www.cs.cmu.edu/~thrun.


Kee-Eung Kim

Get Back
Last modified: Thu Jan 14 12:24:11 EST 1999