The Development of a Versatile Sensor Network for Mobile Robotic Platforms

All mobile robots critically rely on sensor information in order to be able to interact with their environment. These sensors must be connected to the control processor in order for the robot to interpret the information and take appropriate action. However, these sensors often rely on expensive (and often proprietary) hardware in order to communicate with the embedded controller. Often this form of interface does not lend itself well to expansion in the number or heterogeneity of sensor types nor changes in the system controller. This paper outlines the development of a network architecture to support integration of low data rate ranging sensors onto Victoria University´s fleet of mobile robotic platforms. A short-range, 360° field of view sensor network comprising of intelligent sensors has been implemented to test the network architecture. Experimental results are presented giving an indication of the performance of the network.