Analysis of surface strength in frequency domain using fore-aft distance information of wheeled robots on rough terrain

Abstract

This paper explores a possibility of identifying material types using fore-aft distance information of an ultrasonic sensor which is changed depending on surface strength. Fore-aft distance data are transformed into frequency domain using FFT and frequency data are analyzed for extracting features of a material type. Sensor data are acquired by a testbed for analysis of a wheel-terrain interaction on four types of a surface; asphalt, sand, gravel and grass.