Swarm and multi-robot systems have the potential to revolutionize many industries, from last mile delivery to mobility networks. However, debugging these systems, especially when there is a large number of robots involved remains an extremely complex task. To better understand the behavior of these systems, we built a visualization tool capable of collecting relevant information from the robots in real-time and display it onto the robots’ physical working space.
Tangible Swarm builds upon the idea of creating a closed-loop control between a camera that acts as the sensing part (e.g., performing tag tracking, pattern recognition, etc.), an agent-based simulator as the computational element, and a projector that fulfills the actuation component. The proposed tool can be extended to a wide range of use cases ranging from novel robotic art installations to changing urban scenarios.