I. Introduction
The main objective of this paper is to introduce a new hybrid motion planning technique combining potential field approach based on Harmonic Functions (known as harmonic potential fields) and the Probabilistic Roadmaps (PRM). Harmonic potential field is a sub-class of the Potential Field (PF) approach, where the potential satisfies Laplace's equation [1]. The PRM is based on a random sampling scheme to construct an implicit representation of configuration space (C-space) as a roadmap [2]. The efficiency of the PRM in solving the classical robot motion planning problem is reduced in environments where narrow passages exist. In this paper we show that the combination of potential field based on the HF and the PRM can overcome each individual technique's shortcoming in finding a path in environments where narrow passages exist.