Historically, computational interaction with the physical world -- intelligent sensing, actuation, and embodied reasoning -- has required expensive equipment with steep learning curves. Recently, undergraduate resources available for intelligent systems courses have expanded rapidly at prices -- and capabilities -- an order of magnitude lower. What is striking is the lack of options between the extremes of high cost and capability and low cost and capability. By combining and extending current trends in curricula, software, and hardware, this project uses off-the-shelf PCs and the adapts the NSF-supported Python Robotics (Pyro) software system to create a low-cost integration of research-caliber physical agents with common computing resources, refocusing undergraduate intelligent systems courses away from toy problems and platforms toward opportunities for open inquiry. This curriculum weaves units from the PIs' research subfields into courses serving a variety of students across the Claremont Colleges and beyond, to the women of Chatham College and the students of the Community College of Allegheny County. The software and curricular resources developed will be mainstreamed into their existing efforts for widespread distribution. An evaluation team of AI/robotics researchers from Bryn Mawr College, the University of Delaware, and Southern Illinois University will help adapt existing assessment instruments and will provide an external perspective on the results. Their feedback will both serve and measure progress toward the project's fundamental goal: raising the expectations of the power, flexibility, and cost-effectiveness possible in an undergraduate robotics laboratory.
