From CMDash’05 to CMRoboBits : Transitioning Multi-Agent Research with AIBOs to the Classroom

Since 1997, the CORAL research group at Carnegie Mellon, headed by Professor Manuela Veloso, has researched teams of soccer robots using the Sony AIBO robots as the platform (Veloso & Uther 1999; Veloso et al. 2000; Lenser, Bruce, & Veloso 2001a; 2001b; Uther et al. 2002). Our experience runs across several generations of these fourlegged robots and we have met increasing success every year. In the fall of 2003, we created a new course building upon our research experience with the AIBO robots. We have since refined the course and taught it again in 2004 and 2005. The course, which we entitled CMRoboBits: Creating an Intelligent AIBO Robot, introduces students to all the concepts needed to create a complete intelligent robot. We focus on the areas of perception, cognition, and action (illustrated in Figure 1), and use the Sony AIBO robots to help the students understand in depth the issues involved in developing such capabilities in a robot. The course has a two-hour weekly lecture and a one-hour weekly lab session. The coursework consists of weekly homeworks and a larger final project. The homework assignments include written questions about the underlying concepts and algorithms as well as programming tasks for the students to implement on the AIBO robots. Evaluation is based on the students’ written answers, as well as their level of accomplishment on the programming tasks. All course materials, including student solutions to assignments, are made available on the web. Our goal is for our course materials to be used by other universities in their robotics and AI courses. In this paper, we present the list of topics that were covered in the lectures and include examples of homework assignments as well as the rationale behind them. This curriculum shows how the AIBO and its software resources make it possible for students to investigate and embody an unusually broad variety of AI topics within a one-semester course.