Flexible functional role allocation between agents and human team members is a crucial issue in effectively supporting teamwork. The RETSINA agents support all dimensions of team work in multiple ways.

1. Agent-Based Support for Team Situation Assessment
   • act as an information provider
   • ensure relevant knowledge is exchanged among all members
   • provide shared representation of the situation to support common knowledge
   • remind team members if pertinent information is currently ignored
   • tell team members when information source is disabled/unreliable
2. Agent-Based Enhancement of Team Supporting Behaviors
   • make visible others' actions and decisions so that potential errors can be detected
   • recognize deviations to current plans
   • alert members to potential errors, constraints and conflicts
   • support team self-reflection and self-correction
3. Agent-Based support for Team Leadership/initiative
   • communicate intent behind commands to allow lower levels to meet intent under evolving conditions
   • communicate priorities and notify members when priorities change
4. Agent-Based Support of Communication among Team Members
   • alert members to ambiguity
   • support translations of terminology among various subgroups
   • serve as a repository for messages between team members
     • e.g., audio to text so information is not lost
     • support self-awareness and debriefing
     • enable data mining/learning
   Different types of team aiding by agents have been experimentally tested with human subjects.

   
   
   Multidisciplinary University Research Initiative (MURI)
   Principal Investigator: Katia Sycara
   Sponsored by: Office of Naval Research (ONR)
   ONR Contact: Michael Shneier
   © 1998 Carnegie Mellon Robotics Institute