Domestic self-propelled start-up company Tucson has received investment from the global chip giant NVIDIA, after the investment, the British Wei Dazu Tucson future 3% stake. The investment was included Sensor in the future of Tucson financing. Automated driving truck market cake is very large, each company entry point is not the same. And gradually into the aging society will cause the capacity crisis; truck accident rate is much higher than the safety hazards of passenger cars; bad driving habits caused by fuel waste ... ... long-term long-term transport in China, some of the pain points, Tucson future Aiming at this market segment to try to reshape the freight industry through an automated trucking program.
Tucson future CEO Chen Mo has repeatedly said in public that Tucson has no passenger car planning. At the beginning of the establishment, and most of the technology with the start-up companies, Tucson future by choosing a strong field of artificial intelligence technology to determine the company's business direction. But in practice they found that artificial intelligence technology, although there is a certain degree of breakthrough but there are still restrictions. In addition, in different segments, the value of artificial intelligence is also very different. Automatic driving truck solutions can reduce freight costs and improve freight efficiency. From this perspective, business logic is established. The scene Fuel Rail Pressure Sensor of the truck is relatively fixed, most of the high-speed section, the technology is also a subtraction process. In the automatic driving industry chain, start-up companies from the L2, L3 level of automatic driving cut and there is no obvious advantage. There are two modes in the future of Tucson, one is to use the goods belonging to their own assets to complete the transport when there is freight demand; the second is to cooperate with the shipping company to maintain all the technical solutions and bear the corresponding responsibility. Automatic driving is generally divided into perception, positioning, decision-making and control of the four modules. The use of various types of sensors on the vehicle around the vehicle information to achieve accurate perception for the follow-up module to lay the foundation, its importance is self-evident.
Tucson's sensor solution is based on computer vision, with 8-10 cameras with different focal lengths deployed on the truck roof and combined with millimeter-wave radar for data acquisition. As the main driver of automatic driving, the camera to monocular, binocular or even multi-purpose, based on computer vision technology to achieve object ranging, object recognition and other functions. But only rely on the camera's sensor program is flawed, will be extreme weather and light and other factors. The addition of millimeter-wave radar solves this problem. Through data fusion and other technologies from different sensors to analyze the data processing, to achieve complementary advantages to calculate the traffic within the field of vision of the location, speed, trajectory and other related information. This sensor fusion program for the subsequent decision-making, early warning and other functions to provide a good bedding, but also for the perception of automatic driving to provide a more complete solution. Positioning module is an indispensable part of automatic driving technology, according to the collected data in real time to build high-precision three-dimensional map, help to automatically drive the vehicle to a deeper understanding of their own environment. Tucson future plans vehicles equipped with GPS, IMU, camera and laser radar, multi-sensor return to the characteristics of the points to complete the match, the lane line, isolation belt, traffic signs and other information to accurately build, to ensure that the body positioning error of less than 5 cm The Hao Jia male introduction, in view of the future Temperature Sensor of Tucson automatic driving truck application is limited to fixed sections, so the cost of building the map is not high. Tucson's future choice for autonomous drawing of high-precision maps is an important reason is that the data provided by the business does not meet their needs. Hao Jia male said that the current high-precision map field and no ready-made standards, each map of the positioning technology is not the same. Self-study method is Tucson's technical core, in the environment perception, positioning navigation, decision-making control and other automatic driving links are reflected. In the case of decision-making, Tucson uses the algorithm of rule and learning fusion. Rule-based short-range decisions, and learning-based long-range decisions help automating vehicles to determine next actions such as braking, deceleration, and tracking.
The future of Tucson is being tested in a bicycle, and the next step may be tested in units of the team. The advantage of the test team is that the car can follow the car, you can streamline the sensor, while reducing the wind and thus Speed Sensor reduce fuel consumption, cohesion can also be improved. But from a technical point of view, the team's reliability requirements for communication is relatively high.