Development of a Sensory Data User Interface for a Service Robot

Abstract

This paper describes the ongoing development of a sensory data visualization module which is conceived to be integrated into the user interface of a manipulatorequipped service robot for assistance tasks. The purpose of the visualization module is to display to the user meaningful shape approximations of the scene starting from the data provided by proximity sensors available on the tip of the manipulator. This information should allow the user to supervise the robot in performing difficult tasks such as identifying and picking up an object from a high shelf or laying in an area not accessible to navigation. Of the two surface reconstruction algorithms investigated, the Power Crust interpolating algorithm has proven better suited for the sparse and uneven information typically returned by proximity and tactile sensors, whereas Hoppe's approximating algorithm deals more e#ciently with the large data sets generated by range scanners.