We will develop a control system for swarms of Unmanned Air Vehicles (UAV) based on biological principles of self-organization and emergent behavior. We will leverage our company's expertise in the areas of complex systems, swarm intelligence, self-organization and agent-based modeling. UAV swarms rely on local, distributed control and communication strategies of the type observed in insects and other natural groups. This leads to greater robustness than traditional, centralized control approaches, while yielding substantial reductions in code size and communication requirements. These practical advantages make it possible to deploy UAV swarms with shortened development times and at a fraction of the cost of other existing approaches. We will define quantitative metrics to evaluate the effectiveness of the proposed approach and to allow for systematic comparison with other traditional approaches. In Phase II we will extend this work through biologically inspired learning and interactive evolution techniques, and will collaborate with a UAV manufacturer to apply our results to the control of real UAV swarms. Our result will show how to generate desired emergent behaviors through the manipulation of local control and communication rules. The tools we develop for controlling UAV swarms will be applicable to a variety of tasks in both military and private settings. The same principles will also be applicable to other types of swarms, for instance wheeled or legged robots to be deployed in: search-and-rescue missions; inspection and repair in hazardous araeas; surveillance; intelligence gathering.
