Self-explanation capability for cognitive agents on-board of UCAVs to improve cooperation in a manned-unmanned fighter team

摘要

This article proposes the concept of a self-explanation capability as part of an automation concept, which enables the pilot of a single-seat fighter aircraft to manage more than one unmanned combat aerial vehicle (UCAV). The overall automation concept bases on the theory of cognitive and cooperative automation and is described at the beginning. It suggests that unmanned aircraft are equipped with on-board artificial cognitive units (ACUs), by which they are enabled to exhibit cooperative capabilities and rational behavior in the context of the mission. To accomplish efficient man-machine cooperation the need for support of the pilot will be identified. The presented concept proposes to support cooperation by providing explanations about UCAV behavior to the pilot in order to assist the coordination, the development of a mental representation of the situation and the evolution of trust. Explanations can be either requested by the operator or provided on own initiative of the cognitive agents. The presented concept has been realized as laboratory prototype and already been evaluated with operational personnel in our human-in-the-loop full scenario simulation environment in an experimental campaign. The selected experimental setup is described and results are discussed with regard to the impact of the self-explanation capability towards man-machine cooperation.