Autonomous cars navigate in the dark: lights off
The tests were performed on Ford’s Proving Ground in Arizona (USA) as part of the carmaker’s Ford’s Smart Mobility program. This R&D program has been implemented by Ford to drive future-prone technology developments such as connectivity, autonomous driving and digitization of mobility in general.
According to Ford, the night rides mark the next step towards the development of completely autonomous vehicles. They prove that a car can rely entirely on LIDAR technology, doing away with cameras and visible light. Hitherto, the widespread belief was that LIDAR, while offering a superior image resolution in comparison with radar, does not reach the quality of camera images. In connection with a high-resolution digital 3D road map and the virtual driver software, the LIDAR-led vehicle was able to follow even twisting roads safe and precise enough.
The LIDAR sensor in use scans the surroundings up to 2.8 million times a second. The system utilizes infrared light with a range of 200 meters to generate a digital 3D image of the vehicle’s environment. These data are compared with a previously stored digital high-resolution road map containing comprehensive data that describe the road, road markings, geography, topography, and even traffic signs, buildings and trees. The vehicle’s autopilot computer compares the LIDAR signals reflected by the environment with the map and the radar data in realtime and translates possible deviations into steering and braking instructions.
Until some time ago, LIDAR systems were considered as much too expensive for deployment in series vehicles. The distinctive LIDAR sensors Google used in its first robot cars are said to carry a price tag comparable to that of a compact car alone. Recent developments in LIDAR technology however have contributed to significant price and size reductions.
Ford is driving the development towards autonomous vehicles also in Europe: In cooperation with the RWTH technical university of Aachen (Germany), the company is developing innovative HMI concepts which are regarded as an indispensible element in the evolution from driver assistance system to autonomous driving.