A virtual framework of robotic SWARM testbed

Abstract

A robotic SWARM testbed is presented in a companion paper, as complementarities, we present a virtual framework of this robotic SWARM testbed for the same objective. The similarities of these two approaches are both trying to implement formation control algorithms onto robots and generate certain criteria to adapt the algorithms to the dynamics of the agents. Comparing to the physical testbed, this virtual framework has a lot of benefits. For instance, the agents can be either any kind of commercialized robots or even the ones not existing that we can design by our own, which means it could have some advanced features, such as self localization capability and global knowledge of their circumstances. These are demonstrated using an example of coordinated standoff tracking task. Furthermore, a virtual framework enables provision of co-simulation capability which is critical for SWARM testbed