Wikipedia:R-CAST

R-CAST is a group decision support system based on research on naturalistic decision making. Its architecture, based on multiple software agents, supports decision-making teams by anticipating information relevant to their decisions based on a shared mental model about the context of decision making.

Principles of design
In this digital information age, decision-making teams are often flooded with an overwhelming amount of information. This leads to two challenges:
 * First, a human decision maker can be overloaded with information and have difficulty making good decisions in a timely manner.
 * Second, members of a team may have difficulty determining what information a teammate actually needs, and hence what information needs to be shared with him/her.

The R-CAST technology aims to address both of these challenges.

The R-CAST approach is based on four major concepts:
 * 1) Agents use a model of human decision making process (called recognition primed decision [RPD] model) to link decision-making tasks to information relevant to the decisions.
 * 2) The computational RPD model in R-CAST uses a knowledge structure (called experience knowledge) that captures knowledge relevant to decision-making.
 * 3) Three types of relevant information can be anticipated from experience knowledge and inference rules, relating to:
 * 4) matching current situation to known experience (i.e., cues),
 * 5) evaluating multiple decision options, and
 * 6) detecting anomalies after a decision is made so that the original decision can be modified accordingly.
 * 7) The computational RPD model serves as a shared DM process between agents and human in a team, which enables agents to share relevant information to other teammates, whether they are software agents or human.

Principles of operation
In addition to anticipating information needed for decision makings, R-CAST agents also collaborate to seek and fuse information in a distributed environment such as Service-oriented architecture. R-CAST is developed at the Intelligent Agents Laboratory in the College of Information Sciences and Technology at Pennsylvania State University, led by Dr. John Yen.

The R-CAST architecture is component-based and reconfigurable. By selecting components suitable for an application, R-CAST can be configured into a wide range of agents: from simple reflex agents to RPD-enabled agents. Key components of R-CAST include the RPD model interpreter, the knowledge base, the information manager, the process manager, the communication manager, and adapters for various domains. The RPD model interpreter matches the current situation with known experiences, which are organized into a hierarchy. Missing cues relevant to the current decision are identified. The information manager uses the information dependency in the knowledge base to infer missing information that is relevant to the higher-level cues, option evaluations, and anomalies. The communication manager then contact agents that provide the missing information. To build a model, one has to (a) determine what components are involved to compose the model, (b) analyze tasks and elicit relevant knowledge for the components, and (c) develop domain adapter to integrate agents to the external environment.

R-CAST agents have been used to develop decision-making aids for human teams. They have also be used to study team cognition and issues related to human-agent collaboration in time-stressed application domains.

Publications

 * Xiaocong Fan, Bingjun Sun, Shuang Sun, Michael McNeese, and John Yen, RPD-Enabled Agents Teaming with Humans for Multi-Context Decision Making, AAMAS 2006
 * X. Fan, S. Sun, M. McNeese, and J. Yen, Extending Recognition-Primed Decision Model For Human-Agent Collaboration, In Proceedings of the Fourth International Joint Conference on Autonomous Agents and Multi Agent Systems, pp. 945–952, The Netherlands, July 25–29, 2005.
 * X. Fan, S. Sun, B. Sun, G. Airy, M. McNeese, J. Yen, Collaborative RPD-Enabled Agents Assisting the Three-Block Challenge in Command and Control in Complex and Urban Terrain, In Proceedings of 2005 Conference on Behavior Representation in Modeling and Simulation (BRIMS), pp. 113–123, Universal City, CA, May 16–19, 2005.
 * X. Fan, S. Sun, and J. Yen, On Shared Situation Awareness for Supporting Human Decision-Making Teams, In Proceedings of 2005 AAAI Spring Symposium on AI Technologies for Homeland Security, pp. 17–24, Stanford, CA, Mar. 2005.