Gong Mengze. As a high-frequency word in the recent industry, urban NOA (pilot assisted driving) is a high-level function of assisted driving, and it is also the only way to truly unmanned driving.
The first shot of urban NOA was Xiaopeng, which has been deeply engaged in the field of driving assistance, and Xiaopeng is also one of the first car companies to achieve high-speed NOA. As early as September 2022, Xpeng took the lead in pushing the urban NOA function in Guangzhou, and the following year, it released the first phase of its own NOA product XNGP for the first time.
In addition to Xpeng, car companies such as Ideal, Weilai, and Jihu have successively put forward their own urban NOA function visions, named their own exclusive NOA technology names, and successively started fixed-point tests, during which the concept of "Kaicheng" came into being. The so-called Kaesong refers to the number of cities in which the NOA technology of car companies can be applied. The speed of opening the city and the number of cities have become the criteria for judging the maturity of NOA technology and the size of market coverage.
In the author's opinion, there are two reasons why major car companies are piling up to promote urban NOA. First of all, the marketing methods and content of the new energy vehicle market have undergone profound changes. The marketing focus of major manufacturers has shifted from traditional elements such as mileage and interior configuration to more fashionable and technologically charged intelligent driving functions.
Secondly, urban NOA, as the ceiling in assisted driving, is the "cake" that the head pioneer must eat. The leading car companies focusing on intelligent driving have completed the construction of algorithm frameworks, and what is lacking is the collection and filling of data after the implementation of more urban scenarios. Polishing the urban NOA function can be seen as a preparation for the trial of evolving to a higher level of L3 autonomous driving.
It should be pointed out that urban NOA is a new thing, and it will inevitably encounter labor pains in the early stage of development, and there will inevitably be poor driving experience such as inaccurate perception system, insufficient intelligent route planning, and insufficient decisiveness in parallel overtaking. The author believes that it is advisable to start from three aspects: strengthening the R&D strength of intelligent driving, strengthening the insight into user needs, and digging deep into basic software development, so as to realize the upgrade and progress of NOA step by step, so as to lay a solid foundation for entering the era of fully autonomous driving.
First, polish high-quality products and further enhance product maturity and perfection. For car companies, while "rolling" to NOA, they must also take into account the cost of software development. At present, urban NOA is one of the few intelligent configurations that can appear as a paid option. However, with the shortening of the development cycle of new models and the increase in quantity, it is difficult for the sales revenue of a single car to bear the high software cost, which leads to the poor experience of the developed intelligent driving function.
Second, continuous data reinforcement should be carried out to solve the problem of low refinement of the urban road network. In urban operation, NOAs generally have problems such as route breakpoints, obvious hesitation of vehicle AI, and low steering success rate, which indicates that car companies may have the suspicion of one-sided pursuit of the speed of opening the city, and the refinement of the actual road section is far from enough. In the future, the urban road network should be further refined and enriched, and finally the urban NOA can be operated independently with a low rate of manual intervention.
Third, strengthen the research and development of basic software to create intelligent driving products that are satisfactory to users. As a key element supporting NOA, the growth trend of automotive basic software is directly related to the application experience of NOA. As the competition for intelligent driving has entered a white-hot stage, enterprises should dig deep into the targeted development of automotive basic software, on the one hand, to provide users with rich and interesting entertainment and game options to create a unique and differentiated competitive advantage. On the other hand, it can form a good adaptation and integration with localized algorithm and microkernel enterprises, and can carry out low-cost iterative migration of cross-vehicle functions in the future, fully realize the reuse of software assets, effectively reduce costs, and improve technical strength and application maturity.
* |Station cool Hailuo production |Zhang Wenling