Analysis of Camera, Millimeter Wave Radar and Body Controller

In fact, when it comes to ADAS, the first things that come to mind are ACC (adaptive cruise control), LDW (lane departure warning), LKA (lane keeping), and FCW (front vehicle collision warning). These technologies are mature, and the products have been mass-produced. They are equipped on high-end vehicles in recent years. And these technologies are all related to cameras, millimeter-wave radars, and body controllers, and seem to have little to do with maps. In fact, if combined with maps, these functions will become more powerful. Let's give a few examples below.

　　ACC is actually a maximum cruising speed set by the driver. The front sensors on the car, such as cameras or millimeter-wave radars, confirm whether there are vehicles in the passable area directly in front of the vehicle. If there are vehicles, follow the vehicle within the safe braking distance. If there are no vehicles, accelerate to the set cruising speed. Note that the emphasis here should be placed on "directly in front of the vehicle", that is, the straight direction of the front of the car. If you are driving in a straight line, or the road curvature can be almost regarded as a straight line, this adaptive cruise function is great.

　　However, if the vehicle is in a curve, as shown in the figure below, the sensor of the vehicle in the left lane will mistakenly think that the vehicle in the right lane is in front of it, so it will slow down immediately. But in fact, the vehicle can maintain normal speed when turning.

　　Or, as shown in the picture below, the vehicle in the left lane is completely unaware that there is a car in front of it at the turn in its lane. Perhaps when the vehicle follows it, or when the vehicle in front suddenly brakes, it is too late to slow down, and an accident is likely to occur.

　　In addition, the map can provide the curvature of the road, so that the vehicle can slow down in advance when passing a curve to ensure safety. Or when passing a curve, the camera's recognition mode can also be changed to improve the accuracy of detecting lane lines or traffic participants on the curve. The map can also tell when the vehicle will enter and leave the tunnel, and the corresponding camera can adjust the exposure rate in time to ensure the continuity of perception.

　　When driving at night, the vehicle can automatically adjust the angle of the headlights left and right at bends according to the curvature information provided by the map, and can also adjust the angle of the headlights up and down according to the slope information provided by the map.

　　On the other hand, many camera recognition providers currently use cameras to identify speed limit signs on the road to achieve the speed limit function, but in fact maps can provide better services. For example, maps can not only tell vehicles the speed limit on the road section, but also provide recommended speeds on dangerous sections, bridges, tunnels, and on and off ramps. Not only that, it can also be refined to the recommended speed of each lane. It is also necessary to emphasize here that in the case of active safety and autonomous driving, the recommended driving speed is more practical than the road speed limit.

　　Based on the slope data, the vehicle can accelerate appropriately before going uphill, and reduce the throttle when reaching the top of the slope, and do not refuel when going downhill. The Electronic Horizon made by Continental is based on the above, and it is said to save energy by 3%. This function is even more significant in saving daily operating costs for commercial vehicles and logistics companies.

　　There are also relatively basic autonomous driving that can be achieved based on maps. For example, in the scenario of autonomous driving on highways, when driving in a straight line or in a lane, the vehicle relies on the main functions of ADAS, that is, ACC+LKA+FCW. When it is necessary to change lanes, the instruction to change lanes can be a manual turn signal, or it can be a lane change decided by the autonomous driving system itself.

　　At this time, the lane information provided by the map is used to confirm whether the current vehicle area can change lanes. If you are going to exit the ramp, the map can notify the vehicle of the ramp entrance ahead at a long distance in advance. At this time, the vehicle has enough time to use the map data to slowly move to the rightmost lane and prepare to exit the ramp. If you do not exit the ramp, or there is a place where traffic merges, the traffic can also have enough time to move to the leftmost overtaking lane to ensure driving safety.

　　In addition, there are a lot of creative application methods. Next time, we will introduce the technical solution of map implementation in ADAS and the working principle of ADASIS.