Our project, SWARM, is a team of autonomous robots designed to quickly and reliably create a 2D map of a large indoor space such as a warehouse. Rather than relying on a single computer, or “leader robot” to control the entire fleet, each robot will independently process information about its surroundings and communicate with the other robots to build a shared map.
This distributed approach makes the system more resilient because the robots can continue working even if one or more robots temporarily lose their connection or stop functioning. The robots will use sensors such as LiDAR and an IMU to determine where they are and identify the obstacles in the room.
Currently, most warehouses operate by having a designed layout that dictates the organization of the warehouse. While ideally the physical layout should match the design, human error and accidents can cause the layouts to diverge. SWARM is designed for such situations, verifying that layouts match, or catching divergence before it becomes a problem.