Robotic Manipulation
Robot arms working in cluttered, unstructured settings like kitchens or restaurants are the focus of this MIT OpenCourseWare course. It covers perception through deep learning and 3D geometry, planning including robot kinematics, trajectory generation, collision-free motion planning, task-and-motion planning, and planning under uncertainty, and dynamics and control using both model-based and learning-based methods. Homework assignments walk students through building a software stack that lets a robotic arm autonomously manipulate objects in cluttered scenes. A final project lets students dig into a chosen subtopic in depth, with real hardware available for ambitious projects alongside heavy use of cloud-based simulation. Materials come from MIT's electrical engineering and computer science curriculum and are freely available through MIT OpenCourseWare under a Creative Commons license, with no certificate offered.