Object-oriented robotics

Abstract

The real-time control of a robotic vehicle is a highly complex task. The vehicle consists of several components, all of which must be controlled both individually and collectively for the vehicle to operate properly. The control of individual components generally falls under the realm of classical control theory. Coordinating the activities of all the components, however, goes far beyond that. Concepts from artificial intelligence, computer vision, navigation, and graph theory are all used together to implement 'human-like' reasoning for problems such as motion control, path planning, and obstacle avoidance. An object-oriented approach helps to segment the overall control problem into more manageable subproblems, which should lead to more dependable code.