Distributed Formation Control of Autonomous Vehicles

Research Impact

In recent years, robotic platforms have persistently become more modular, miniaturized, and affordable, while at the same time their onboard computational and communication capabilities have advanced significantly. Thanks to these trends, it is now possible to deploy a large number of robots to cooperatively and autonomously execute tasks such as environmental mapping and monitoring, infrastructure inspection, delivery of goods, and intelligent transportation. Distributed control of robotic networks facilitates a large-scale deployment of robots, which is naturally parallelized, resilient to loss and failure of communication and hardware, and suitable for preserving information privacy.

Our Contribution

Our research in this area has centered on cooperative formation control of autonomous vehicles. Each vehicle in the group senses, or receives via communication, certain relative position and orientation information about a limited set of other vehicles. Based on this limited information, local control laws are designed such that a prescribed geometric configuration emerge from the interaction of vehicles, while ensuring a safe behavior such as collision avoidance. Our work has revealed the relation between the structural information architecture and the stability properties of the desired configuration under different vehicle dynamics and control laws. We have tested our formation control strategies on simulation platforms such as Matlab and AirSim, and further experimentally implemented them on several multi-robot testbeds including Sphero robots, Crazyflie quadrotors, and Robotarium.

Applications and Simulation Results

Our proposed formation control strategy can be used for vehicles with a variety of dynamics such as quadrotos:

differential-drive robots:

and cars: