Adaptive Levels of Automation for UAV Supervisory Control

Problem UAVs hold great potential for force multiplication and increased functionality. To deliver on this potential, future UAV systems must include reliable and flexible autonomous control. UAV autonomy is limited. The desireable level of autonomy depends on many factors. Ideally, the level is dynamic, raising and lowering in response to the mission phase and realworld contingencies. This project will develop an Adaptive Level of Autonomy (ALoA). The system itself continuously determines the appropriate balance of control. Approach 1. Begin with existing successful mulit-UAV simulation. 2. Articulate components, define autonomy levels. 3. Develop variable levels of autonomy. 4. Define cooperative UAV protocol. 5. Develop adaptive mechanisms. 6. Create human interface. 7. Test and iterate. Result Phase I results in a functioning ALoA design. This design manifests itself in three distinct deliverables: 1. A sophisticated realtime simulation demonstrates the ALoA system in operation. The simulator offers supervisory control of several overlapping missions. To each mission is assigned a single UAV or a cooperative UAV group. A variety of different UAV profiles are modelled in the simulator. 2. A whitepaper reports on the research and declares a set of ALoA principles. 3. A sophisticated DVD summarizes Phase I in presentation-oriented format.