Ongoing research on multi-robot collaborative navigation conducted in the Intelligent Robotics Lab at Vanderbilt University is introduced. Range-free egocentric representations for navigation have been previously developed for a single robot. We refer to this method as perception-hosed navigation (PBN). This paper will show that the concept of PBN can be extended from a single robot to a group of robots to achieve collaborative navigation. One robot attempts to reach a target region by sharing its sensory information with the knowledge of another. PBN approach separates sensing and navigation from planning. Our assumption is that one robot is lost or new in an environment, which is known to another robot. The new robot asks for assistance by presenting its egocentric world perceptions. We show the PBN idea is expandable to handle collaborative efforts to achieve a goal.
展开▼
