Support of teamwork in human-agent teams

Our research program is investigating how to effectively incorporate intelligent agents into human teams, specifically to support team performance. There are several ways to incorporate agents into teams; our approach is to use agents to support the team as a whole (facilitating communication, allocation of tasks, coordination among the human agents, and improving attention focus). In our initial experiments, we investigated the effects of trust and calibration of trust that allows human decision-makers to assess the reliability and meaning of communications from software and human agents. Currently, we are focusing on the team and the different ways of deploying agents to support multiperson teams: 1) supporting the individual by tracking the collected information; 2) supporting communication among team members by automatically passing information to the relevant person; and 3) supporting task prioritization and coordination by providing a shared checklist. Only performance data have been reported for this paper. Aiding with agents was found to be different from the control condition. Learning effects and effects for the target difficulty were found