GRASP Home Page

Home | Goals | People  Research | Publications | Robot Platforms | Multimedia | Links | Grant

Multi-Robot Localization and Mapping

John Spletzer and C.J. Taylor


Simultaneous Localization and Mapping (SLAM) poses a tremendous challenge to a single robot. However, sharing information amongst a team of robots can greatly simplify the task.
Using a team of Clodbusters (CBs), a global map of the environment can be generated up to a scale factor. With a minimum team size of 3 robots, localization of each in 3D space is possible. This is accomplished by using azimuth and elevation information obtained from the CB omnicams. With this relative information from each platform, localization for each can be accomplished using triangulation. Each team member can then augment a portion of the global map with its local map information.

A summary briefing on our initial exercise - SLAM with a 3-robot team and using a ground plane constraint - can be found here. Mapping results will follow soon.



Home | Goals | People  Research | Publications | Robot Platforms | Multimedia | Links | Grant