Three-dimensional discrimination between planes, corners and edges using ultrasonic sensors

This paper presents an ultrasonic sensor made up of several transducers located on various different horizontal planes, designed for navigation and localization applications in the mobile robot (MR) field. The sensor proposed is equipped with two emitters and fourteen receivers. The times-of-flight (OTF) provided by the transducers enable the tree types of basic reflectors (planes, corners and edges) to be classified. Optimal classification of the reflectors will facilitate both localization of the mobile robot and the generation of environment maps that are useful for navigation tasks. The innovation with this sensor as regards prior papers in this field lies in the high degree of redundancy that it provides, which greatly facilitates the detection of erroneous measurements caused by false reflections, and in the simplicity of the classification algorithm used, which gives it a low computational load.