Surfing the Sensorium : Navigation on Sensory-Motor Trajectories

Autonomous navigation of a mobile robot along a predefined trajectory is a widely studied problem in the robotics community. As an alternative to Cartesian trajectories, we propose to define trajectories as sequences of sensory-motor information. In this paper we show that such a representation is suitable for navigating on the trajectory. To this end, we describe an architecture of Bayesian inference to localize and control the robot on its trajectory. In order to increase robustness, we also use this Bayesian framework to estimate system self-confidence while the robot is moving. This work has been validated both on a simulated robot and on a real car-like robot: the CyCab.