Autonomous vehicles in structured and semi-structured environments

Abstract

We consider real-time decision-making issues regarding the mobility of different classes of autonomous mobile platforms. We discuss the implications of different environments and the development of a unifying control architecture. Such an architecture would make the parameterization of behavior easier, and introduce the possibility of a comparison of intelligence. We report on our development of different autonomous systems working in a coordinated fashion in three different environments: automobiles and automated highway systems, model railroads, and wheeled mobile robots. We identify five basic tasks which are common to these different environments, and we discuss their interpretation and implementation in each environment.