Search for a command to run...
In this article, we have demonstrated, through the parking and leaving example, how high-level specifications containing multiple temporally dependent goals can be given to a team of realistic robots, which in turn automatically satisfy them. By switching between low-level feedback control policies and moving in a well-behaved environment, the correctness of each robot's behavior is guaranteed by the automaton. The system satisfies the high-level specification without needing to plan the low-level motions in configuration space. Sensor inputs play a crucial role in this framework. A hazard input becoming true at the wrong time may lead to deadlock. Deciding when and how long to stop is a hard problem even for humans, as sometimes demonstrated at four-way stops, let alone robots.
Published in: IEEE Robotics & Automation Magazine
Volume 15, Issue 1, pp. 30-38