ABPEM—Agent-Based Planning, Execution and Monitoring

ATC implemented an agent architecture called Agent-Based Planning, Execution, and Monitoring (ABPEM) that utilizes a dynamically reconfigurable, data-flow network for planning, execution, and monitoring on large-scale multi-information-source systems. This approach is applicable to a broad range of systems such as transportation, logistics, manufacturing, and information gathering, especially where the domain is characterized by dynamic environments that change during the course of execution. It was applied to the problem of validating and verifying airlift cargo schedules. Given a schedule of airlift cargo missions, relevant data is identified and collected and then used to (a) validate the future scheduled missions, and (b) verify the completion of executing missions (or reporting unexpected results).

This research will support critical military C4I applications as well as other planning applications in autonomous unmanned vehicles and battlefield robotics. Civilian applications include intelligent highway and air traffic control; medical robotic assistance; work-cell manufacturing; industrial inspection; autonomous farming; nuclear facility maintenance; personal assistants (softbots); and mobility aids for the handicapped.