Task-based interfaces for decentralized multiple unmanned vehicle control
Enhanced autonomy in Unmanned Vehicles (UV) has given human operators the ability to move from teleoperation to supervisory control of single vehicles, and now multi-vehicle coordination. This research seeks to leverage task-based interfaces, where the human operator guides a fleet of decentralized UVs via high-level goals as opposed to individual vehicle control. In such decentralized control architectures, each vehicle computes its locally best plan to accomplish the mission goals with shared information. The results of two experiments are described where 62 participants performed multi-UV missions in an existing decentralized multiple unmanned vehicle simulation environment under increasing task load. Results suggest that a system which uses a task-based interface and decentralized control algorithms may be robust to task load increases by mitigating operator cognitive overload.
