Real-time collision-free motion planning in a non-stationary environment is an important issue in many autonomous systems including robotics and intelligent systems. It endows intelligent robotic systems with an ability to plan motions and to navigate autonomously. This ability becomes critical particularly for robots which operate in dynamic environments, where unpredictable and sudden changes may occur. This project addresses the fundamental issues associated with robot path planning in a dynamically changing environment subject to uncertainties. This book presents a methodology based on a unique combination of neural network techniques and dynamics, and cellular automata. Cellular Neural Networks (CNN) take the best of both path planning methodologies, in that autonomous navigation, conducted in real-time within a dynamic environment can be successfully performed. By harnessing the elegant properties of harmonic functions, and the nonlinear local interconnectivity of neurones within a CNN, the complete elimination of local minima can be obtained producing a stable and convergent path planning methodology.