Autonomous, Ground Robotic Vehicles For Electrical Engineering Technology

This paper discusses the design and provides an analysis of the performance of an autonomous ground robotic vehicle (GRV) called Vasilius. Three sections of focus are presented. First, the paper describes the hardware and software design of Vasilius. A novel idea of modeling an autonomous vehicle after human senses and the human decision-making process is presented. For example, Vasilius integrates a large number of independent sensors such as stereoscopic vision to incorporate short-range reaction as well as long-range planning. In addition, the GRV uses two computers in a decentralized control scheme. Secondly, the paper reports the performance of Vasilius, relating theoretical predictions to actual behavior. Performance measures include object avoidance, lane following, waypoint accuracy, follow-the-leader, ramp climbing, turn reaction time, battery life, stop reaction time, behavior in dead-ends and traps, pothole avoidance, and safety. Finally, the paper discusses current on-going research topics with Vasilius, including the incorporation of Kalman Filtering, mapping, and learning. The important side benefit of implementing this applied research robotics project in the ELET 492 Senior Design course, where student teams build robotics for our defense agency clients, is realized with their success in the annual International Ground Robotics Vehicle Competition. This program has proven to be a definite benefit to motivating our students as well as satisfying our clients.