Communications for Agent-Based Human Team Support

AbstrAct This chapter discusses the problem of agent aiding of ad-hoc, decentralized human teams so as to improve team performance on time-stressed group tasks. To see how human teams rise to the challenge, we analyze the communication patterns of teams performing a collaborative search task that recreates some of the cognitive difficulties faced by teams during search and rescue operations. Our experiments show that the communication patterns of successful decentralized ad-hoc teams performing a version of the task that requires tight coordination differ both from the teams that are less successful at task completion and from teams performing a loosely coupled version of the same task. We conclude by discussing: (1) what lessons can be derived, from observing humans, to facilitate the development of agents to support ad-hoc, decentralized teams, and (2) where can intelligent agents be inserted into human teams to improve the humans' performance.