Domestic LiDAR is gradually overcoming sensor hardware, fusion algorithms, and introducing roadside data, and China is expected to overtake on curves.
Regardless of whether the car company promotes L2.9 or L3+, the number of sensors installed in cars is indeed increasing.
The fifth generation auto drive system of Waymo, the global leader of unmanned driving, has 40 sensors, including 29 cameras, 6 millimeter wave radars and 5 laser radars. The new forces in domestic car manufacturing are not inferior in terms of automotive intelligence and automation. The NIO ES8 is equipped with 23 sensors, while the Xiaopeng P7 has 31 sensors, surpassing Tesla.
With the arrival of the intelligent transformation of automobiles, global automotive technology companies, big and small, are committed to removing the steering wheel of cars or partially reducing the role of human drivers. Some players are scaling new heights and advancing towards L4 level autonomous driving, while others are gradually starting from L1 and L2 level autonomous driving. However, the market cake for in vehicle sensor solutions has been growing bigger and bigger.
The "2020 Autonomous Driving Sensor Report" released by market research company Yole D é evelopment predicts that sensors used for autonomous vehicles will grow at a compound annual growth rate of 51% over the next 15 years, and the total revenue of sensing hardware will reach $17 billion, approximately RMB 111.5 billion, by 2032.
In the billion dollar market for automotive sensors, China has lost the lead in research and market share. The automotive sensor industry, such as cameras, millimeter wave radar, and LiDAR, has been monopolized by foreign brands for many years. However, some domestic brands are eager to regain the dominance of domestic sensors.
How can cameras, millimeter wave radar, and LiDAR make cars perceive the environment like humans? Can image recognition alone enable cars to reach the level of fully autonomous driving? Beyond the end of the car, how will the end of road sensors help autonomous vehicle land faster?
The "eyes" and "ears" of a car
Compared with traditional cars, the main work of the perception system of autonomous vehicle is to replace the human driver's vision system: by converting the perceived information into electrical signals according to specific rules, it is transmitted to the central control unit of the car to assist in automatic driving.
Autonomous driving ultimately aims to make cars intelligent, and there is often a saying in the industry about how artificial intelligence can help humans: the areas that current artificial intelligence systems excel in are exactly the opposite of those that humans excel in. Human logical analysis ability far exceeds AI, but humans are far behind AI in areas such as memory and big data analysis.
The same phenomenon exists in car perception systems. Human drivers only need to rely on common sense and cognition to identify pedestrians, vehicles, traffic lights, etc. on the road, and make corresponding decisions such as acceleration, deceleration, and turning. However, it is very difficult for computers to perform the same operation, and perception and recognition are the first major obstacles.
At present, the core sensors of automatic driving include vehicle camera, millimeter wave radar, laser range finder sensor and laser radar. In ADAS system, camera and millimeter wave radar are mainly used, and laser radar has become a must for most autonomous vehicle above L3 level.