A harmonic potential field approach with a probabilistic space descriptor for planning in non-divisible environments

This paper extends the capabilities of the harmonic potential field approach to planning to cover the situation where the workspace of a robot cannot be segmented into geometrical subregions each having an attribute of its own. Instead the suggested planner accepts a holistic, task-centered, probabilistic descriptor of the workspace as an input. This descriptor is processed along with a goal point to yield the navigation policy used to direct motion. The extension is based on the physical analogy with an electric current flowing in a nonhomogeneous conducting medium. Proofs of the ability of the modified approach to avoid zero-probability (definite threat) regions and converge to the goal are provided. The capabilities of the suggested planner are demonstrated using simulation.