Multi-agent systems have been employed to study distributed control and collaborative tasks using intelligent autonomous agents are capable of sensing, acting, cognition and communication. Distributed control systems have the advantage of producing solutions that are more robust, adaptive and fault tolerant since there is no critical reliance on any specific individual agent, and do not require the full pre-specification of problem solutions. This project will focus on the role of explicit symbolic communication between agents in distributed control tasks. This can be crucial in tasks requiring higher level cognitive capabilities, such as planning and decision making in distributed control scenarios. This will also support the reconciliation of human decision making schemes with machine performance and intelligent agents. The project is based on a series of simulation experiments on collaborative tasks such as search and rescue scenarios in unknown and unpredictable environments. In particular the focus will be on the emergence of symbolic communication lexicons from the interaction between autonomous agents. A comprehensive analysis of the results in the various simulation experiments will produce a theoretical framework for communication in distributed-control domains. This will provide meaningful insights on the optimal distributed-control strategies for the design and improvement of human-centered distributed control systems.

People

Prof. Angelo Cangelosi (University of Plymouth)

Prof. Leonid Perlovsky (Harvard University and US Air Force Research Laboratory, Hanscom AFB)

Mr. Fabio Ruini (University of Plymouth)

Related links

Centre for Interactive Intelligent Systems

Adaptive Behaviour and Cognition Research Group

Disclaimer

The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either expressed or implied, of the Air Force Office of Scientific Research or the U.S. Government.