On April 19, 2022 HAOMO AI DAY was officially held in Beijing. The theme of the event was "Accelerate to the Future". Chairman Zhang Kai and CEO Gu Weihao shared the latest strategies, technologies and product progress of Feimo, and announced a number of achievements and major announcements: Feimo will firmly rank among the mass producers in China in 2022 Number one in autonomous driving. In terms of passenger cars, the mileage of assisted driving users has exceeded 7 million kilometers. China's first large-scale mass-produced urban assisted driving product "Haimo City NOH" will be officially launched in the near future; in the field of terminal logistics automatic distribution vehicles, Haimou Xiaomotuo 2.0 Officially released, single The vehicle sells for 128,800 yuan, making it China’s first 100,000 yuan terminal logistics automatic distribution vehicle; in terms of technology, China’s first data intelligence system MANA has received a number of major upgrades, and its perception capabilities have improved by leaps and bounds. Through the original “dual-stream” perception Models and self-developed BEV Transfomer, "makes China without unrecognizable traffic lights and lane lines" a reality; cognitive capabilities, costs and evolutionary speed have also doubled, model training costs have been reduced by 60%, and the acceleration ratio has exceeded 96%. In addition, the AI automation rate has been marked Reaching 80%, significantly reducing labeling costs.
With the continuous increase in the penetration rate of high-level intelligent driving in passenger cars, the NOH intelligent pilot assisted driving system has gradually entered the city, and intelligent driving in urban scenes is becoming a A must-win for new military strategists. Starting from 2021, Haimo has released two generations of passenger car assisted driving HPilot products, which are being delivered in an orderly manner as planned. At this HAOMO AI DAY, Haomo launched the "Haomo City NOH" equipped with HPilot3.0. This is China's first large-scale mass-produced urban assisted driving product, the first urban assisted driving solution that emphasizes perception, and is also China's first most practical and efficient urban assisted driving product in 2022.
At the scene coverage level, the system can automatically change lanes in an urban environment based on the driving route provided by the navigation It has main functions such as overtaking, traffic light recognition and control, complex intersections, and unprotected left and right turns. It can also handle complex urban traffic scenarios such as vehicles cutting in at close range, vehicles blocking the road, intersections, roundabouts, tunnels, overpasses, etc.
At the hardware level, the system is equipped with An ultra-high computing power chip with an AI capability of 360T, a cache of 144M, and a CPU computing power of 200K+ DMIPS; it is also equipped with 2 lidars, 12 cameras, and 5 millimeter wave radars to maximize the complete set of assisted driving perception systems.system security redundancy.
In terms of product strength, the intersection passing rate of passenger cars equipped with this system exceeds 70%, the lane change success rate exceeds 90%, and the traffic flow processing capability is as high as level 4. It is understood that in daily tests, the intersection pass rate of "Haomo City NOH" is ahead of Tesla's performance in China.
At present, "Haomo City NOH" has been undergoing in-depth scene polishing in cities such as Beijing and Baoding, and vehicles equipped with this system will also be officially mass-produced in the near future. In the future, the "Haomo City NOH" function will be implemented in more than 100 cities, and the number of passenger cars installed at the same time will exceed 1 million.
In the follow-up interview session, Bitauto also proposed the city version of Haomo. Can it be successfully completed in the case of extreme weather or emergencies? And how difficult is it to achieve urban NOH at night compared with during the day? How did Hao Mo overcome it? problem. Hao Mo gave the following answer:
Hao Mo: To answer your first question, first of all, the current driving systems are mainly assistive. , it is assisted automatic driving or assisted driving. It is impossible to cope with it like a human being, or humans cannot cope with all situations, and there will be system limitations. There are indeed some situations that the system cannot handle, such as extreme close-range congestion, or some tiny obstacles that suddenly pop up, such as a cat or a dog. Our system will iterate quickly. For example, based on MANA's cognition, when starting, if there is a big bus on the right, but it does not start at the traffic light, or it suddenly starts and blocks your view. At that time, I can stop starting, or I can stop in front of its stop line, and tentatively see what happens next to it. Because at that time we were rapidly iterating and responding in MANA based on human driving experience or human driving behavior. This is the answer to your first question.
The second question is that the scene at night is actually far more complicated than during the day, because we have high beams when we drive at night. Lights, some very dazzling traffic lights, including searchlights, and the blurring and contrast of routes at night are some of the difficulties we have in identifying and dealing with them. Different regions have different levels of wear and tear, including some differences in our road pavements. And this is the basic reason why we are vigorously promoting MANA now, because we want to use the entire set of MANA capabilities to quickly solve the industry's problems. Only by doing this can we quickly push away the NOH in the city.
At the end of the event, Gu Weihao said: "HaiMo is an AI technology company dedicated to autonomous driving. Technology is the unchanging belief of BiMo people. Everything in the future of BiMoPossibly, it all comes from the daily technical hard work of the people at Bimo. Technological improvement is the tireless pursuit of our people, who strive to improve every moment. ”