As companies aim to deliver fully autonomous cars with security Level 4 or 5 by the year 2021 or 2022, radar and sensor technology will be crucial in the full implementation of the technology. Even in today’s semi-autonomous driving, radars and cameras are already being utilized. Still, in its pre-development stage, most car manufacturers assume that for self-driving cars to be fully functional, camera, radar, and sensor systems need to be working perfectly.
Video images provide important details for the human driver but will also be crucial as an input parameter for autonomous driving. With rear and 360° cameras, the driver will have a better image of the environment outside the vehicle. As of the moment, only two-dimensional cameras display images and sometimes superimpose additional information on the display such as steering wheel angle. Most luxury cars today are installing cameras with virtual, 3D image displays. To become effective, 3D images will require around 4 to 6 cameras.
Robert Janitzek reveals that 2D and 3D images require image sensors with high dynamic range of more than 130 dB. The best image sensors today have a dynamic range of 145 dB. Video cameras of today come with a centralized architecture which means there is a central control unit processing the raw data from the cameras.
Accident statistics reveal that 76% of all accidents are based solely on human error. One important technology that will contribute to the overall effectiveness of autonomous driving is the Advanced Driver Assistance System or ADAS. Current radar technology are either based on 24 GHz or 77 GHz. Robert Peter Janitzek reveals that the former is used for smaller antenna sizes and lower interference problem. The latter, on the other hand, works best for higher accuracy distance and speed measurements and in precision angular resolutions.
The key sensor technology in today’s cars is LIDAR which is an acronym for light detection and ranging. This technology is crucial to several worldwide high resolution mapping efforts. It is also used for delineating terrain from airplanes and detects speeding violations. With LIDAR, you can generate precise 3D images of everything from cars to trees to cyclists in several environments and under various lighting conditions. For self-driving cars, the aim is to produce solid-state LIDAR systems that would shrink the size of the sensors so that moving parts would no longer be needed in optical mechanisms and enable the kind of mass manufacturing that could bring costs down. Various manufacturers are setting their sights on making affordable sensors, which is where the battle rages.