MokSAF imagemap

Research Objectives
We are developing simulated environments in order to study how mission critical teams make decisions and negotiate resources. In such cases where humans must cooperate in order to satisfy team goals, software agents act as intelligent actors by sorting and evaluating enormous quantities of information. Incorporating software agents into human teams, however, raises several questions. For example, should agents be considered members of the team or personal assistants? What role(s) should agents play in overall decision-making? Should an agent's role change in relation to team performance and task level? How can we ensure that agents will be able to anticipate dynamic situations and objectives?

In the context of joint mission planning, we are interested in how human teams put our two main agent types (i.e. planner agent and critique agent) to work in order to plan a route and make equipment choices that satisfy both individual and team constraints. While the Path Planner Agent takes account of physical constraints when planning a route automatically, it can not account on its own for social or dynamic constraints (e.g. areas such as hospitals and schools that commanders may wish to avoid). The Critique Agent allows team members to participate actively in the construction of all or part of a route, but it cannot suggest a shorter or more cost efficient route. The Critique Agent can only say whether or not the user's suggested route will work, given what it knows about the terrain's physical and social constraints. Thus, we are interested in how team members negotiate the costs and benefits of different agent types.

Applied to mission critical teams, agent technology has the potential to:

In sum, our research is interested in how teams, when faced with time pressures and dynamic environments, can work together successfully.
Major research questions include:

What roles should agents play in decision-making?

How do team members deploy different agent types?

How will agent technology impact team decision making?

Home - Overview - Objectives - Agent System - Contact Us - Publications - People - Demo - Internal Site (Restricted Access)

We welcome your comments and suggestions.
Copyright © 1999

Carnegie Mellon University
Software Agents Group
Robotics Institute
   University of Pittsburgh
Dept of Information Science
& Telecommunications
Project Overview Research Objectives Description of MokSAF Agents Contact Information Publications Researchers and Support Agencies System Demonstration Internal Site MokSAF Home Page