Fuzzy constraint satisfaction approach for landmark recognition in mobile robotics

Abstract

Landmark detection is essential for topological navigation, and many type of sensors have been used for this task. Ultrasound (US) sensors are one of the most common in mobile robots, but their readings include a high degree of imprecision and noise, which makes it hard obtain reliable detections with them. In this work, we apply a model, based on fuzzy constraint satisfaction problems (FCSP), for detecting doors over US sensors in mobile robots. Experimental results using a Nomad 200 mobile robot show 91% of doors correctly detected, which supports the reliability and robustness of the proposed methodology. 1