Recently, Zhang Peng, CFO of Momenta, was a guest in “The People’s Living Room” and discussed the theme of “Autonomous Driving in the’Second Half’ of Conspiring on New Trends in Future Travel”. At the same time, at the “Xiandao AI to” Xiangdao Robotaxi Online Experience Day held in Jiading, Shanghai, Momenta officially announced that it will provide Xiangdao Robotaxi with an L4 autonomous driving solution, which will compete with SAIC Passenger Cars, SAIC Artificial Intelligence Laboratory, and Xiangdao Robotaxi. Travel together to explore the application and innovation of L4 autonomous driving technology in urban travel scenarios.
It is understood that in Xiangdao Robotaxi’s L4 autonomous driving solution, Momenta’s “flywheel L4” technology is fully applied to enhance Robotaxi’s continuous evolution capabilities and create a more scalable autonomous driving: The process “data-driven” algorithm, compared with the conventional “rule-driven”, achieves a cross-order improvement in the performance of each module; through the Momenta Framework, the data flow and technical flow can be opened up, and the L4 technology is provided with “mass data” support; through Momenta, it is efficient The “closed-loop automation” tool chain of the company can carry out highly automated mining and highly automated labeling of golden data, and realize efficient and automated iteration of massive data processing.
Zhang Peng believes that the two major applications for L4 and L5 products will be Robotaxi and Robotruck. The common point is that they can ensure safety and save labor costs on a large scale. The social value and economic benefits brought by L4 autonomous driving will be reflected in these large-scale applications in the future. He said that if the first half is “0-1” and the second half is “1-N”, the biggest test here is the ability to scale, which is where Momenta’s unique “flywheel” advantage lies.
“There are two difficulties in realizing this application, one is the technical level, and the other is the supporting policy level. Technically, in the face of the million-level long tail problem, a large amount of inefficient manual labor must be abandoned. Momenta is driven by data Method, to build a system that automatically discovers and solves problems, called “a flywheel”, is also the technical insight at the beginning of the company’s establishment.” Zhang Peng said that data-driven is the engine, massive data is the fuel, and closed-loop automation is the accelerator. “Flywheel” will be able to efficiently iterate algorithms for massive long-tail problem scenarios, realize large-scale unmanned driving more efficiently, and empower safer, more convenient and efficient future smart travel.
“At the policy level, at the level of L4-L5, the responsibility of driving is shifted from human drivers to cars. This change will bring about a series of legal problems. It requires supporting laws and regulations in various related fields such as roads and transportation to support autonomous driving. We are pleased to see that China’s policy is very forward-looking. On November 25, Beijing’s Smart Connected Vehicles Policy Pilot District issued the Implementation Rules for the Pilot Management of Automated Pilots (for Trial Implementation). The policy states that Robotaxi operators can provide services for a fee in the pilot area. We believe that this marks a new stage in the domestic autonomous driving field from testing and demonstration to commercial piloting. It is important for accelerating autonomous driving technology Landing applications to change the way of travel in the future is of great significance.” He said.
Regarding the two modes of bicycle intelligence and vehicle-road collaboration in the process of autonomous driving, Zhang Peng said that bicycle intelligence relies on the vehicle itself to sense, predict, and plan operations. The discovery and resolution of massive long-tail problems is also Momenta’s “a flywheel.” The focus of strategy is at the core. The vehicle-road collaboration is the organic combination of “people-vehicle-road-cloud” transportation elements through the Internet of Vehicles. The purpose of the two methods of bicycle intelligence and vehicle-road collaboration is to ensure the safety of autonomous driving. Many basic technologies including vision and perception are the same. At the hardware level, a large number of sensors, radars, etc. are also required, and there are many commonalities.
“Momenta will actively participate in the research and development of vehicle-road collaboration. We announced a strategic cooperation with China Mobile last year, combining Momenta’s leading full-stack capabilities of autonomous vehicles, roads, maps, and clouds, and conducted a small-scale pilot in Suzhou. In the future, Momenta will continue to pay attention and actively participate, so that’smart cars’ are equipped with’smart roads’.” Zhang Peng said.
Regarding the vision of future travel, Zhang Peng mentioned that autonomous driving technology will free humans from mechanical driving tasks and no longer need to sit behind the steering wheel. The car of the future becomes a mobile office, a mobile cinema, or a mobile gym, etc. Make full use of the time on the road to greatly improve the efficiency and pleasure of travel. It is believed that with the help of “flywheel L4”, the future of large-scale unmanned driving can be expected. Also just as our mission describes “Better AI, Better Life”.