Coordination and simulation of multiple robots in the game of hockey

For the past decade, modeling and control of multi-robot systems have been one major area of robotics research. Major advantages of a multi-robot system include its ability to solve complex problems that cannot be achieved by a single robot, and its robustness in the event of robot malfunction. Promising testbeds for multi-robot system research include robot soccer and hockey, in which a team of robots coordinate their actions for goal-scoring in a dynamic, adversarial environment. This thesis presents the framework for design, development, and simulation of multi-robot coordination in the context of robot hockey. Potential field construction and super-position are performed to determine strategic waypoints for shooting, passing, and dribbling. An auction is performed to dynamically assign actions to each robot in a team. Progress monitoring is used to coordinate actions between human-controlled and computer-controlled robots. The software simulator features: robot kinematics, event dynamics, game strategy testing, and result visualization.