Is stereo vision a suitable remote sensing approach for motorcycle safety? An analysis of LIDAR, RADAR, and machine vision technologies subjected to the dynamics of a tilting vehicle.

Tilting vehicles, such as electric bicycles, motorcycles, and scooters, are increasing in popularity as a means of personal transport. From a safety viewpoint, the development of Advanced Rider Assistance Systems (ARAS) for two-wheeled vehicles is lagging behind the Advanced Driver Assistance Systems (ADAS) for other road vehicles (e.g. autonomous emergency braking implemented for passenger cars and trucks). This study is the first analysis of three remote sensing technologies adopted by ADAS, such as RADAR, LIDAR and machine vision, but from a point of view significantly different to the used in the car industry. Essentially, the dynamics study of a four wheelers vehicle cannot be used because it does not take into account a tilting dynamics. Our findings indicate that the lack of technology transfer from ADAS to ARAS can be explained by sensor design considerations, which limiting the application of the existing automotive remote sensing approaches to tilting vehicles. We propose how to tackle these limitations in terms of both hardware and software, by presenting related experiments on a scooter.