Abstract:

This report presents details of the aim, the preliminary results and the proposals for future development of the research project entitled “Multisensor Synergy for Understanding a Robot’s Environment”. The aim of this project is to develop a new multisensor synergy system in order to investigate the challenges and the benefits of developing synergistic sets of sensors for understanding a robot’s environment. Currently algorithm to integrate data from disparate sensors (i.e. temperature, humidity, ambient light intensity, compass bearing, air velocity and direction, roll and pitch) is being developed using the Basic Leader-Follower clustering technique. Early results show that the algorithm can autonomously group readings of disparate sensors in a specific environment. The groups created represent the average measurements of sensor readings at some particular locations within a period of time. Future work will consists of development of a technique to map groups of disparate sensor readings based on their spatial and temporal distribution. This map will be used by a mobile robot to become aware of the normal features of its surrounding environment. By knowing the normal states of the environment, the robot can recognize and adapt to any new events through some investigation using the multisensor system. The development of the ‘investigation’ algorithm will also be the major part of the future work.

Speaker Biography