Researchers from Caltech have developed a machine-learning algorithm which allows drone swarms to navigate in tight spaces without bumping or hitting into each other. This algorithm, called Global-to-Local Safe Autonomy Synthesis (GLAS) could be useful in fancy drone light shows, as well as vital operations, such as search and rescue operations.
Instead of relying on existing maps or the routes of every other drone in the swarm, GLAS has each machine learning how to navigate a given space on its own even as it coordinates with others. This decentralized model both helps the drones improvise and makes scaling the swarm easier, as the computing is spread across many robots.
An additional tracking controller, Neural-Swarm, helps the drones compensate for aerodynamic interactions, such as the downwash from a robot flying overhead.
Wow!
(Image Credit: Aerospace Robotics and Control at Caltech/ YouTube)