Flexibility models for robot navigation

In this paper we propose a uniied mathematical framework for local robot navigation and position localization. This framework is based on the use of exibility maps of the environment. Flexibility maps contain both the current knowledge of the robot about the positions of the objects in its environment as well as the information required to calculate its future path. The major beneet of our approach is that several navigation tasks in mobile robotics can be incorporated elegantly into a single mathematically well founded framework. Also, we found that the simple control law derived from this framework generates paths that look very natural in realistic circumstances, adding to the plausibility of the approach. This control law requires angular and/or distance information about a set of landmarks relative to the robot which can be provided easily by simple vision and/or range sensors.