A Novel Approach to Road Scanning for Automotive Simulations 2020-01-5039

Road scanning techniques are currently used worldwide to assess and manage the road infrastructure. In the last 15 years, the same techniques and technologies have been used to improve vehicle simulations and are now being used to drive the development toward autonomous vehicles. The industry-wide push for virtual engineering requires new methods and tools to be developed in order to capture the reality with higher resolution and accuracy. This paper presents an innovative approach to road scanning for automotive simulations, hereby defined as “autonomous terrestrial laser scanning.” It combines a mix of autonomous elements, the accuracy of terrestrial laser scanning (TLS), and submillimetric resolutions given by image stereoscopy techniques. A potential implementation for this approach is introduced: A robotic platform is used to integrate multiple optical sensors with the objective of measuring surface wavelengths from 100 m down to 25 μm. The custom navigation algorithm drives the robotic platform along a predefined plan and stops at specific locations to perform measurements. In this way, maximum accuracy is delivered by each sensor with no need for inertial correction as in standard mobile laser scanning (MLS) systems. The results show how this novel approach can deliver surface geometric properties that can be directly used in vehicle simulations as well as support the study of rubber-asphalt interaction.