Using Agents and Robotic Soccer for Undergraduate Education

Using Agents and Robotic Soccer for Undergraduate Education

  Robotic soccer is a well-known challenge problem for artificial intelligence and robotics. Two well-known annual competitions, RoboCup and FIRA allow teams to compete in a number of different leagues distinguished by robot size and hardware restrictions. We have been using RoboCup soccer as a means of teaching applied artificial intelligence to undergraduates, and have found it to be a very motivating vehicle for encouraging undergraduate exploration in this area, as well as an effective means for teaching students about artificial intelligence, real-time systems, and software engineering. It is also, however, a domain whose complexity must be carefully managed in order to allow undergraduates to participate successfully. This paper describes our strategies for using robotic soccer for undergraduate education, and in particular the agent-based approach that we employ to make the domain manageable for students at this level.