Current status of China's intelligent driving NOA technology development

Yiou released the "2024 China NOA Functional Technology Development Roadmap Insight Analysis Report", and made some summary of China's current intelligent driving NOA.

We have discussed the market size, technology route, competitive landscape, cost control, and future challenges. Based on this report, we also discuss that the penetration rate of China's NOA (Navigation on Autopilot) function will reach nearly 10% in 2024. The demand for intelligent driving technology is increasing, especially in high-speed driving and urban scenarios. The application of intelligent driving functions will become more extensive.

01

The development of NOA functions is divided into two parts: high-speed scenarios and urban scenarios.

● Urban NOA has become a new high ground for technological competition, especially among leading companies. In order to compete for the market, they have accelerated the introduction of NOA solutions that do not require high-precision maps and can cover national applications.

As the cost of technology decreases, the NOA function will be gradually installed in mid- and low-end models. It is expected that the price of models equipped with high-speed NOA will drop below 150,000 yuan in the future, especially Baojun and Aion will take the lead. The smart driving NOA function is not only exclusive to high-end users, ordinary consumers will also have the opportunity to experience this technology, further promoting the popularization and growth of the market.

In the development of NOA technology, major automakers and suppliers have chosen different technical routes, which is particularly evident in the selection of perception systems and algorithms.

The report points out that the core differentiation of NOA technology currently lies in the differences in perception configuration and algorithms.

Some companies have chosen a multi-sensor fusion solution that combines multiple sensing devices such as lidar, millimeter-wave radar, and cameras, while other companies advocate a pure visual route. Companies represented by Tesla have vigorously promoted the development of pure vision technology in the development of urban NOA functions, removing the dependence on high-precision maps, reducing costs while improving the versatility of the system.

02

NOA's LiDAR "arms race" has ended. Chinese auto companies have turned to more cost-effective solutions based on actual conditions, with forward-looking binocular cameras and multi-sensor fusion solutions becoming mainstream choices.

The total cost of the forward-looking binocular camera solution is about RMB 2,000-2,500, while the cost of the multi-sensor fusion solution is RMB 5,000-6,000. This differentiated technology choice provides more flexibility for models with different market positioning.

The choice between multi-sensor fusion and pure vision is not static. High-end models still tend to adopt the multi-sensor fusion route, while mid- and low-end models will gradually try pure vision solutions. This differentiation of technical paths determines that car companies will adopt different NOA development strategies for different user groups and different market positioning in the next few years.

With the gradual development of NOA technology, cost control has become one of the key factors. The cost of high-speed NOA solutions can currently be controlled within 10,000 yuan, and is expected to further drop to 5,000 yuan in the future. The annual reduction in costs will not only promote the popularization of NOA functions, but also promote the penetration of NOA technology into mid- and low-end models.

● New entrants into the high-speed NOA function usually choose lower-cost pure vision solutions, competing with traditional multi-sensor solutions through cost-effective technical routes.

Compared with solutions that rely on high-precision maps, map-free or map-light solutions can significantly reduce development and operation costs, which provides great advantages for the promotion of NOA technology in urban scenarios. The application of emerging technologies such as multimodal large models, cloud simulation and digital twins has enriched the simulation test scenarios of NOA and further accelerated the commercialization process of urban NOA.

At the competitive level, urban NOA has become a new focus of corporate competition in 2024. The old leading car companies are accelerating the layout of full-scenario NOA, while new entrants are rapidly advancing the implementation of urban NOA functions by memorizing driving modes. The competition in intelligent driving technology has gradually expanded from a competition of single technical solutions to a comprehensive competition for market share, cost control and the speed of technology implementation.

For companies providing intelligent driving solutions, it has become a trend to expand market share through differentiated customized services, and they are increasingly focusing on rapid application of technology through cooperation with vehicle manufacturers to shorten development cycles and increase market share. As the complexity of intelligent driving technology increases, full-stack self-development has become a key way to achieve independent control, and cooperation is the fastest path to commercialization.

At present, NOA technology faces problems such as heavy rules, poor generalization, and high cost. The complex roads and traffic environments in urban scenarios place higher demands on perception, decision-making, and control systems.

The safety and reliability of NOA technology remain the focus of industry development. How to ensure the stable operation of the system and reduce misjudgments and accidents will become an important direction for technological breakthroughs in the next few years. NOA technology also needs to achieve further progress in algorithm optimization, system redundancy and test verification.

summary