Application of ultrasonic guided waves to robotic occupancy grid mapping

This paper evaluates the benefits of using ultrasonic guided waves for the mapping of a structure, when implemented on a mobile magnetic robotic platform. It considers the specific problem of mapping geometric features using the guided ultrasonic waves, which enables the localisation of edges and/or the welded joints. Shear Horizontal (SH) guided waves generated by Electro-Magnetic Acoustic Transducers (EMATs) are used for mapping steel samples with a nominal thickness of 10 mm. A Bayesian mapping technique (Occupancy grid mapping) was used to map the boundaries of an irregular sample in a pseudo-pulse-echo mode. The principle is demonstrated in both simulation and laboratory-based experiments. It is shown that the proposed mapping algorithm successfully estimates the position of a sample's edges. Experimentally, a range accuracy of < 1.7 mm (1σ) was achieved on a 1 × 2 m sample using miniaturised EMATs operating at a wavelength of 22 mm.