12:00, 23 October 1996, WeH 7220


		Landmark-Based Mobile Robot Navigation

			   Sebastian Thrun

To operate successfully in indoor environments, mobile robots must be
able to localize themselves. Over the past few years, localization
based on landmarks has become increasingly popular.  Virtually all
existing approaches to landmark-based navigation, however, rely on the
human designer to decide what constitutes appropriate landmarks.

While you guys are all eating, I will chat about some new research
that enables our mobile robot AMELIA to select its landmarks by
itself.  Landmarks are "discovered" by training neural networks so as
to minimize the a posteriori localization error that the robot is
expected to make after querying its sensors.  An empirical study
illustrates that self-selected landmarks are superior to landmarks
selected by a human (=Tom).