Intelligent Agent Integration of On-Board and Off-Board Sensors Systems for Mobile Robot Path Planning

The paper describes a novel approach of using distributed intelligent agents for the integration on-board and off-board sensors for mobile robot path planning in indoor environments. The system gives the robot a prior knowledge about the path state, which will be taken by the robot during navigating from an initial position to a target point. This enables the robot to intelligently decide in advance which path should be taken before reaching a point where a barrier to progress may have appeared. The system takes advantage of surveillance cameras already in buildings to aid the robot to plan a path. We have built intelligent agent based system that supports controlling the robot efficiently, using cameras available in the building and managing the communication between agents. The multi-agent system architecture includes two platforms, each platform consist of different agents connected together wirelessly using a JADE middleware platform.