Multi-robot Systems Lab (MSL)


In the Multi-robot Systems Lab (MSL) we study strategies for controlling groups of robots to accomplish useful tasks in a distributed fashion. Some examples include controlling a team of autonomous helicopters to provide surveillance of an environment, or deploying a group of
autonomous rescue vehicles to locate and retrieve survivors in a collapsed building. The key difficulty in controlling such systems is the lack of a centralized coordinating station; the robots must determine their actions in a distributed fashion using the limited information available to them. We use mathematical tools from control theory, optimization, and information theory to design and analyze algorithms for multi-robot systems to perform coordinated control, sensing, and learning tasks. To validate our algorithms, we rely on mathematical proof, verification in computer simulations, and hardware experiments with robotic platforms. Ultimately, we seek to produce distributed multi-robot behaviors that are as sophisticated and robust as the distributed phenomena that pervade the natural world, such as birds flying in a flock, colonies of termites building a complex nest structure, or groups of cells cooperating to make an organism.