An Intelligent Agent for Supervisory Control of Teams of Unihabited Combat Air Vehicles (UC.pdf

Presented at the Unmanned Systems 2000 Conference, Orlando, FL, July 11-13, 2000. 1 An Intelligent Agent for Supervisory Control of Teams of Unihabited Combat Air Vehicles (UCAVs) Dr. Mark L. Hanson, Karen A. Harper Charles River Analytics 725 Concord Ave. Cambridge, MA 02138 (617) 491-3474 (617) 868-0780 (fax) mhanson@ Achieving air superiority is a key tenet of U.S. fighting doctrine. To meet this requirement, flight packages of Uninhabited Combat Air Vehicles (UCAVs) have recently been proposed for future conflicts. One operations concept calls for UCAVs to possess significant onboard intelligence that enables real-time decision-making without continuous human intervention. In this case, although the vehicles will be highly autonomous, risk mitigation and other considerations still require humans-in-the-loop for UCAV operation. Because the operator will be supervising multiple vehicles, advanced decision aids are needed to maintain accurate situation awareness. This paper describes a decision aid called OLIPSA (On-Line Intelligent Processor for Situation Awareness), which was originally developed for advanced fighter aircraft cockpit displays, but has recently been extended to supervisory control of UCAVs performing Suppression of Enemy Air Defense (SEAD) missions. 1. Introduction In the 1991 Persian Gulf War and more recently in Kosovo, hundreds of sorties were devoted to locating and destroying enemy air-to-ground defenses. Because these missions require planes to fly into the heart of Integrated Air Defense Systems (IADS), there is a great risk for loss of pilot life. To reduce this risk, as well as to more rapidly respond to Presented at the Unmanned Systems 2000 Conference, Orlando, FL, July 11-13, 2000. 2 changing situations, future U.S. fighting doctrine will use composite strike packages that contain highly intelligent, autonomous UCAVs and traditional piloted aircraft. Though humans will not directly fly these