Motion Planning with Environment Uncertainty
Authors/Creators
Description
As robots are becoming more pervasive and are increasingly used in close proximity to humans and other objects in factories, living spaces, elderly care, and robotic surgery, planning for collision free trajectories in real-time is imperative for safe and efficient operation. Yet, in the presence of noisy sensors, both the robot and world state cannot be estimated precisely. In this paper we investigate the problem of motion planning under environment uncertainty. To this end, we first incorporate object uncertainties in Belief Space Planning (BSP) and derive the resulting Bayes filter in terms of the Extended Kalman Filter (EKF) update equations. We further formulate the collision constraint as a quadratic form in random variables, thus computing an exact expression for collision probability. We further validate our approach using numerical integration and provide a comparison to other approaches.