Aiding Collaboration Among Humans and Complex Software Agents

This paper describes an implemented software prototype for the Distributed Collaboration and Interaction (DCI) system, which addresses the challenges of helping humans to act as an integrated part of a multi -agent system. Human interaction with agents who act autonomously most of the time, such as a process control agent in a power plant, has received little attention compared to human interaction with agents who provide a direct service to humans, such as information retrie val. This paper describes how liaison agents within the DCI system can support human interaction with agents that are not human -centric by design but must be supervised by or coordinated with humans. Further, the DCI prototype supports notification and p lanning for humans from the perspective of an organization. It treats humans in this organization as if they were agents with explicitly modeled roles and activities related to software agents in the same system. Planning for humans presents unique chall enges because the models represented in traditional planners do not match well with human mental models. The DCI prototype is applied in the domain of NASA advanced life support systems, which are controlled by software agents in an autonomous fashion with occasional human intervention.

[1]  Chris Elsaesser,et al.  An architecture for adversarial planning , 1990, IEEE Trans. Syst. Man Cybern..

[2]  David Kortenkamp,et al.  Intelligent Control of Life Support for Space Missions , 2002, IEEE Intell. Syst..

[3]  Mary Beth Hudson,et al.  Automation in Context: Planning Manned Space Exploration Activities , 2001 .

[4]  Cheryl E. Martin,et al.  Specifying Organizational Policies and Individual Preferences for Human-Software Interaction , 2002 .

[5]  Anoop Gupta,et al.  Sideshow: Providing Peripheral Awareness of Important Information , 2001 .

[6]  David Kortenkamp,et al.  Intelligent Control of a Water-Recovery System: Three Years in the Trenches , 2003, AI Mag..

[7]  James F. Allen,et al.  TRIPS: An Integrated Intelligent Problem-Solving Assistant , 1998, AAAI/IAAI.

[8]  Daniel Ryan,et al.  Intelligent Control of Life Support Systems for Space Habitats , 1998, AAAI/IAAI.

[9]  David P. Miller,et al.  Experiences with an architecture for intelligent, reactive agents , 1995, J. Exp. Theor. Artif. Intell..

[10]  Chris Schmandt,et al.  Everywhere messaging , 1999, IBM Syst. J..

[11]  David Kortenkamp,et al.  Adjustable Autonomy for Human-Centered Autonomous Systems on Mars , 1998 .

[12]  Terry R. Payne,et al.  Varying the user interaction within multi-agent systems , 2000, AGENTS '00.

[13]  David E. Wilkins,et al.  A Multiagent Planning Architecture , 1998, AIPS.

[14]  Milind Tambe,et al.  Adjustable autonomy in real-world multi-agent environments , 2001, AGENTS '01.

[15]  Jean Oh,et al.  Electric Elves: Applying Agent Technology to Support Human Organizations , 2001, IAAI.

[16]  John Levine,et al.  O-Plan: A Web-Based AI Planning Agent , 2000, AAAI/IAAI.

[17]  Jean Oh,et al.  Mixed-initiative, multi-source information assistants , 2001, WWW '01.

[18]  Erann Gat,et al.  Experiences with an architecture for intelligent, reactive agents , 1997, J. Exp. Theor. Artif. Intell..

[19]  Peter A. Jarvis,et al.  PASSAT: A User-centric Planning Framework , 2002 .