Modular Robots: Self-Reconfigurable Systems

This lecture explores the perception, control, and planning for self-reconfigurable modular robots. It discusses the advantages of modular robots, the conceptual design of Roombots, search-based planning strategies, search space formulation for motion planning, and the challenges of searching faster. The lecture also covers the division of complex tasks, shape representation with voxels, and the generation of reconfiguration steps. It concludes with insights on user interfaces, computational models for lamprey swimming, and the use of event-based cameras. The presentation highlights the need for novel user interfaces and the potential to recreate animal locomotion using modular robots.