Using Mobile Robots To Explore Intelligent Behaviors: The Obstacle Course Challenge

Abstract

A recently concluded NSF-ILI grant provided equipment to create hands-on laboratories for CIS students. The goal of this laboratory environment was to provide a setting to reinforce course concepts. One of the target courses was the Artificial Intelligence (AI) course. After the AI course, many students desired to pursue more tangible and applied experiences. Additional equipment was purchased to design and construct multiple mobile robotic platforms. With this equipment and a motivated group of students, a special topics course was created. The goal of this new course was to expand on the basic concepts of the AI course. Special topics involving motion, sensoring, planning, fault recovery and hardware/software interfacing were discussed and supported by supplemental readings from the literature. Each topic was linked to a laboratory exercise where students would design and implement both the hardware and software to accomplish an intelligent behavior. The ultimate challenge was to integrate the behaviors from the exercises into the ability to navigate an unknown obstacle course. The obstacle course was designed to include path following, landmark recognition, hitting a target, obstacle detection, obstacle avoidance, and path reacquisition. Students competed in teams of two and each team was evaluated on their robot’s ability to independently accomplish each task. The most successful teams were awarded prizes and shown on the local news report. The instructor found that the AI course served as a solid foundation for this new course and provided the opportunity to explore more advanced topics of AI and robotics. The students, having prior experience in the AI course, could be expected to have a more mature understanding of the literature and use it to develop their robotic projects. This paper will outline the goals of the course, the approach taken, problems encountered and solutions used along with lessons learned.