Enabling Human Support of Robot Swarms

We have developed a framework that, through intuitive visual queues, enables a non-expert user to monitor and support an autonomous robot collective in the presence of errors. Robot swarms with distributed control have great potential. However, they are also hard to debug, monitor, and maintain because the state space grows exponentially with the number of robots, and because errors can propagate in non-intuitive ways.

The framework can help monitor the state of the robots, detect component malfunctions and distinguish robots that behave differently compared to the collective. We believe these decision support systems will be imperative to help translate autonomous robot collectives out of the lab and into real-world complex task settings.