Autopilot must use lidar. If lidar can effectively control costs, driver assistance functions with lower ADAS levels also need to use green laser pointer radar. The reason is that the limitations of camera-based ADAS and driverless systems, or the use of millimeter waves alone, are considerable.
The first is the problem of the field of view. In order to ensure a sufficient detection distance, the angle of the field of view cannot be too large, which results in a vehicle with a very large lateral blind zone. With wide-angle lenses or fisheyes, everyone knows that edge distortion is very serious. Some manufacturers have introduced multi-camera working modes. Even so, it cannot solve the problem of close-range blind spots. The same multi-cameras have overlapping areas. Will increase the complexity of the algorithm, the use of filters and then processing will increase the system response time and increase costs.
The second problem is the low speed. The camera does not perform well at low speeds, and it even fails to recognize those stationary and slowly moving targets. The millimeter wave may only be able to identify obstacles, and it seems incompetent to see whether it is a person or an obstacle. Without lidar, or even looking for alternatives to green laser pointer radar, this road will be difficult.