Low Cost 3D Volume Road Roughness Estimation Model Using Ultrasonic Sensors

Efficiency and road maintenance activities automation possibility (pothole backfilling and bumps leveling) depend on efficacy of disruption volume approximation models. This paper presents a low cost road disruption volume estimation approach as an alternative to existing costly methods (water displacement and computer vision). The proposed method collects data using an array of ultrasonic sensors, applies Chebyshev nodes and Lagrange interpolation to build disruption functions then uses average approximation to slice the disruption and compute its volume. The proposed model is able to suitably estimate placements of n sensors and compute resulted polynomial functions for each measurement at cost of Θ(n2). Twelve potholes and bumps were sampled and captured using ten ultrasonic sensors for test. The method sliced and estimated their volumes with a very high accuracy and precision. Hence, the proposed method is low cost financially and computationally but also practical for both paved and unpaved roads.