Agent-based resource management for smart robotic sensors

We propose an agent-based architecture for managing the resources of robotic intelligent sensor agents (R-ISAs) (Petriu, E. et al., 2002). The main idea is to share all the available sensors among the entire society of agents, in such a way that, even though some of the agents do not have the required physical sensor or actuator on-board, they can always use other agents' resources to overcome this deficiency. The main goal of the project is to allow a human being from a computer station to interact remotely with a society of autonomous robotic sensor agents. The interaction is done through querying and sending commands. The novelty is that the user request leads to an intelligent proactive behavior performed by the agent society. The communication protocols between the agents have been successfully implemented and tested. The development of the Sensor Explorer is underway. The approach we chose is motivated by the success of multi-agent based systems, peer-to-peer (P2P) computing (http://www.openp2p.com), and the flexibility of grid computing (http://www.gridcomputing.com).

[1]  Charles M. Macal,et al.  Action Selection and Individuation in Agent Based Modelling , 2003 .

[2]  Adam Dunkels,et al.  Minimal TCP/IP implementation with proxy support , 2001 .

[3]  E.M. Petriu,et al.  Intelligent robotic sensor agents for environment monitoring , 2002, 2002 IEEE International Symposium on Virtual and Intelligent Measurement Systems (IEEE Cat. No.02EX545).

[4]  Jean J. Labrosse,et al.  MicroC/OS-II: The Real Time Kernel , 1998 .