Liangdao Intelligent launches the first pure solid-state Flash side-viewing LiDAR in the Chinese market, which will be mass-produced in the second half of 2023

On May 13, Liangdao Intelligent, a domestic lidar supplier, officially released its self-developed pure solid-state Flash lateral lidar, LDSense Satellite, to the Chinese market. This is a pure chip design, an automotive-grade lidar product with both performance and cost advantages. It is mainly used for lateral blindness compensation and is scheduled to be mass-produced in the second half of 2023.

According to the scanning method, the technical evolution route of LiDAR can be divided into three stages: mechanical to semi-solid and then to pure solid. It is generally believed in the industry that pure solid-state LiDAR without any moving parts is the mainstream technical direction in the future. Pure solid-state is mainly divided into Flash, OPA and other routes. OPA technology has demanding process requirements and is difficult to produce. It is expected that it will be difficult to achieve large-scale mass production in the short term; Flash LiDAR with relatively simple structure and more mature technology will be implemented earlier. However, the Flash technology route also has shortcomings, that is, low power density and short detection distance, so it is more suitable for close-range blind spot filling. Judging from the LiDARs that have been released, the performance considered by most products lies in the detection distance, and the vertical field of view angle is often very low, usually within 30°, which often forms a large blind spot when detecting at close range.

Flash LiDAR can make up for this shortcoming. LDSense Satellite's vertical field of view can reach 75°-90°. This will play a huge role in some specific scenarios. Full-scenario intelligent assisted driving, including the three major vehicle use scenarios of highways, cities, and parking, is one of the current core R&D investments of car companies. At present, the main scenarios for the implementation of high-level assisted driving capabilities are highways and parking scenarios. The road conditions in urban scenes are complex and the uncertainty of traffic participants is high. It is more difficult to follow, overtake, and avoid obstacles. Most companies are still in the testing stage. Whether the intelligent assisted driving experience is smooth tests the vehicle's ability to identify at close range, which is also where the blind spot LiDAR comes in handy.

In daily driving, both the driver and the vehicle need to pay attention to the road conditions in different driving scenarios at all times, and respond to driving behaviors based on the environmental information captured by the eyes or vehicle sensors. However, both the driver and the vehicle sensor have certain blind spots in vision perception. In complex scenarios such as TJP (traffic jam autonomous driving ) following, highway congestion, narrow road traffic, and autonomous parking, the demand for close-range detection around the vehicle body is particularly strong.

When driving on highways or in urban areas, emergency avoidance scenarios caused by vehicles cutting in often occur. If you fail to avoid in time, it is very easy to get scratched. This requires that the vehicle perceives the lane-changing intention of vehicles in the adjacent lanes as early as possible, and makes accurate judgments and decisions in time to avoid danger. The current mainstream forward-facing laser radar configuration solution in the market generally has a horizontal field of view angle of 90°-120°. When the forward-facing laser radar detects a vehicle cutting in, the distance between the two vehicles is already quite close, and the decision response time is insufficient, so it is easy to collide. Combined with the lateral laser radar, the vehicle's forward perception horizontal field of view can be expanded to 180° or more, and even achieve all-round coverage of the vehicle body without blind spots. When the other vehicle attempts to cross the vehicle, the intention to cut in can be judged, thereby obtaining more sufficient warning and response time, greatly improving the response speed, effectively reducing the occurrence of accidents, and expanding the application scope of L3 and above autonomous driving in the future.

When driving on urban roads, pedestrians, two-wheeled vehicles, and vehicles that shuttle and change lanes are mixed, and the traffic scenes are complex and changeable. In addition, when vehicles parked on the roadside use the intelligent driving function to start and park, they also need to effectively perceive low objects such as the edge of the road and speed bumps. The smaller the height of the obstacle, the more difficult it is to detect, and the higher the requirements for the vertical field of view and angular resolution of the lidar. The forward-facing long-range lidar installed on the roof has a large near-field blind spot during detection and cannot cope with some special scenarios. For example, pet dogs and cats that suddenly jump out in front of the car cannot be captured and detected within 7 meters by the long-range lidar with a vertical field of view of 25° on the roof.

Liangdao Intelligence said that the side-viewing LiDAR with a vertical field of view of no less than 75° can complement the near-field blind area of ​​the forward-viewing LiDAR with point clouds, and can detect low objects such as road edges, speed bumps, and adjacent lane lines. Taking the road edge as an example, at a height of 10cm, it can not only achieve near-field detection of 0.35m-15m, but also achieve ultra-high resolution detection of 30 lines of point clouds at 0.35m.

When the lateral lidar is installed at the top position, in addition to realizing the perception function of targets in adjacent lanes, it can also obtain accurate relative position information by sensing static markers around the road, and combine it with high-precision maps to achieve lane-level precise positioning.

According to reports, LDSense Satellite is the first pure solid-state LiDAR in the Chinese market. Its internal structure uses electronic scanning Flash technology with a pure chip design. The internal structure is simple and has no moving parts. Compared with other LiDAR technology designs, it is easier to meet strict automotive regulations. In the product design and production process, it has ultra-high reliability, longer service life and stronger cost competitiveness.

In addition, LDSense Satellite's compact product structure makes it the smallest product in its class. While meeting the functional definition requirements of different OEMs (original equipment manufacturers), it can be flexibly embedded in the vehicle body and support a variety of integrated solution designs. LDSense Satellite's ultra-large vertical field of view can achieve maximum coverage of near-field blind spots, respond to various driving scenarios in a timely and rapid manner, and can accurately detect even small objects such as speed bumps, greatly enhancing the vehicle's perception capabilities.

At present, there are great differences in the judgment of car companies on the deployment of LiDAR. Not long ago, the CEOs of several new car companies had a debate on this issue. The pure solid-state Flash LiDAR released by Liangdao Intelligence is a differentiated development route. With the implementation of the pure solid-state Flash LiDAR, there will be a more complete sensor configuration solution to prepare for high-level autonomous driving.