Introduction: Urban NOA is Reshaping the Development Path of Autonomous Driving
The past decade has witnessed the progression of autonomous driving industry through stages of technological exploration, scenario validation, and commercial pilot programs. Now it is gradually entering a new era of large-scale deployment. At the early stage, the industry has considered L4 autonomous driving applications such as Robotaxi as a critical direction for commercialization. However, constrained by factors including operational costs, safety validation, and regulatory frameworks, widespread adoption will still require time. Meanwhile, Mass-Production autonomous driving systems represented by Urban NOA have developed rapidly and are gradually becoming one viable pathway for commercial deployment of autonomous driving technology. As technologies keep improving, hardware costs keep falling and consumer awareness continues to rise, Urban NOA has evolved from a differentiated feature for premium vehicles into a core capability of intelligent vehicles, ushering a new development phase for the autonomous driving industry.
Against this backdrop, CIC releases the “Autonomous Driving Industry Blue Paper: Urban NOA as an Inflection Point for Autonomous Driving Commercialization”. Based on extensive industry research and data analysis, this report systematically examines the development trends and competitive landscape of the global and China autonomous driving markets. It analyzes the core reasons why Urban NOA has become an industry turning point, the key drivers behind the development of Mass-Production autonomous driving in China, and the evolution pathways of the Mass-Production model versus the Robo model. Furthermore, by incorporating case studies of representative industry participants, the report explores the strategic value of Urban NOA in advancing autonomous driving commercialization and the journey toward L4 autonomous driving, providing insights for industry participants to understand the future direction of the sector.
Autonomous driving is entering a new phase where Urban NOA will become a key turning point
According to levels of vehicle automation, intelligent driving technology can be categorized into multiple levels ranging from L0 to L4 and above. Among them, L0-L2 primarily fall under the category of Advanced Driver Assistance Systems (ADAS), where intelligent vehicles are equipped with driver assistance functions, while the driver remains responsible for actual driving and continuous supervision. As automation capabilities improve, levels L2+ through L4 and above are classified as Autonomous Driving (AD), featuring stronger environmental perception, decision-making and planning, and driving automation capabilities.
After years of development, the autonomous driving industry is transitioning from the technology validation phase to the stage of large-scale commercialization. In terms of business models, the current mainstream pathways can be categorized into two types: Mass-Production and Robo models. The Mass-Production model refers to the large-scale delivery of vehicles to individual consumers, with commercialization achieved through the integration of autonomous driving functions into Mass-Production vehicles. At present, the most Mass-Production autonomous driving in the industry is primarily concentrated in L2++ Urban NOA capabilities. In contrast, the Robo model mainly refers to L4 autonomous vehicles represented by Robotaxis, which achieve commercialization by providing driverless mobility services within designated operating areas. However, a series of applications under the Robo model, constrained by factors such as stringent regulatory and safety supervision frameworks, face short-term pressure to achieve large-scale adoption. In contrast, the Mass-Production model represented by Urban NOA is becoming the main trajectory of current industry development.
The advantage of the Mass-Production model lies in its ability to continuously amass real-world data through large-scale vehicle sales, forming a virtuous cycle of "data accumulation → algorithm optimization → product iteration." Each mass-produced vehicle equipped with an AD system serves as both a commercial product and a data collection terminal, generating revenue while continuously contributing training data and safety validation samples. As the size of these vehicles grows, the system can more efficiently identify complex long-tail scenarios, enhance model generalization capabilities and engineering maturity, thereby further strengthening product competitiveness.
Against this backdrop, Urban NOA is becoming a critical turning point in the autonomous driving industry. The significance of Urban NOA lies not in the addition of a specific driving function, but in how it transforms the product form and competitive landscape of autonomous driving. In the past, intelligent driving was perceived by users largely as a checklist of features: whether the vehicle is equipped with adaptive cruise control, lane keep assist, automated parking, or automated lane changing. Each function was relatively independent, and users evaluated whether individual functions were usable.
Urban NOA is different from that. It integrates navigation, perception, prediction, planning, control, and human-machine interaction into a continuous driving system. Users no longer focus on whether a vehicle can perform a single automated lane change. Instead, they care about whether the vehicle can naturally follow other cars, change lanes, yield to traffic, pass intersections, respond to traffic lights and avoid roadside parked vehicles, while maintaining safety, comfort and efficiency during the entire urban trip. In other words, the evaluation criteria for intelligent driving are shifting from "whether the feature exists" to "whether the system is trustworthy."
This shift is likened to the transition from feature phones to smartphones. In the feature phone era, competition revolved around a list of functions such as calling, texting, camera, music, and so on. In the smartphone era, the core of competition shifted to the operating system, application ecosystem, user interaction experience, and continuous upgrade capability. Intelligent driving is undergoing a similar migration: traditional ADAS corresponds to the "feature phone," while Urban NOA represents a key step toward "smartphone-like" intelligence, as it integrates multiple functions into a high-frequency, upgradable, and continuous learning system.
Global and China’s autonomous driving markets will maintain robust growth, with Urban NOA contributing significantly
As technology gradually matures and commercialization pathways become increasingly clear, the global Mass-Production autonomous driving market is maintaining robust growth, with its size growing from approximately USD 16.5 billion in 2025 to USD 70.3 billion in 2030. Among them, the China market performs particularly prominently, with its size growing from approximately USD 10.4 billion to USD 38.6 billion over the same period, achieving a CAGR of approximately 29.9%. It will not only become the world's largest autonomous driving market but also serve as a core engine for technological innovation and commercial deployment.
Among various autonomous driving application scenarios, Urban NOA is becoming the most important growth engine at the current stage. Vehicle sales in China equipped with Urban NOA functionality will grow from 2.6 million units in 2025 to 18.2 million units in 2030, with a CAGR of 47.8%; during the same period, the penetration rate of Urban NOA in China will rapidly increase from 11.3% to 62.4%, signaling the transition from a high-end optional feature to the stage of large-scale adoption.
Drivers and trends of Urban NOA development in China
Factors such as OEMs' increasing demand for intelligent differentiation, the data advantages derived from complex road scenarios, an efficient domestic supply chain ecosystem, and growing consumer demand for intelligent driving features have collectively positioned China as one of the leading markets for the development of urban NOA globally. Looking ahead, China's urban NOA industry is expected to maintain strong growth, driven primarily by the following factors and trends:
Policy and supply chain synergy forms a development foundation: In recent years, China has continuously advanced the construction of intelligent connected vehicle testing demonstration zones and gradually improved relevant policy and regulatory frameworks. Meanwhile, a mature new energy vehicle supply chain, a leading intelligent hardware supply system, and abundant road scenario collectively provide favorable conditions for the R&D and commercial deployment of autonomous driving technology.
New energy vehicle proliferation provides a scalable platform: New energy vehicles feature more advanced electrical and electronic (E/E) architectures, stronger computing power, and OTA upgrade capabilities, making them critical platforms for autonomous driving functions such as Urban NOA. As the world's largest new energy vehicle market, China provides ample deployment infrastructure and application scenarios for autonomous driving technology.
Consumer demand upgrade and OEM competition jointly drive the rapid adoption of intelligent driving: As the new energy vehicle market gradually matures, consumer focus is shifting from driving range and power performance to intelligent driving experiences. Urban NOA is gradually becoming an important factor influencing vehicle purchasing decisions. Meanwhile, in China’s highly competitive automotive market, an increasing number of OEMs are leveraging intelligent driving capabilities as a key lever for product differentiation and brand competitiveness, continuously enhancing product competitiveness through rapid iteration and large-scale deployment, thereby accelerating the widespread adoption of Mass-Production autonomous driving technology.
Complex traffic environments drive the continuous evolution of Mass-Production autonomous driving technology in China: The dense mix of non-motorized vehicles and pedestrians, complex intersections, and mixed traffic scenarios on China's urban roads provide intelligent driving systems with rich and challenging training and validation environments, while also raising the industry’s technical bar. Systems that have undergone continuous iteration and validation in China’s complex scenarios are expected to achieve stronger generalization capabilities and global adaptability.
The increasing value of software shifts industry competitions from hardware to algorithm capabilities: As Urban NOA and higher-level autonomous driving gradually become more widespread, the value share of software in autonomous driving systems continues to rise, gradually becoming the core factor determining system performance and user experience. Meanwhile, hardware configurations such as SoCs, sensors, and positioning/navigation are becoming increasingly standardized, shifting the industry’s competitive focus from hardware specifications to software capabilities. Algorithm performance, iteration efficiency, and cross-platform adaptability are emerging as key factors in industry competition.
The share of independent solution providers in the industry continues to rise: Against the backdrop of rapid adoption of autonomous driving technology, an increasing number of OEMs are choosing to adopt solutions from independent providers. Compared to in-house development, leveraging independent providers enables faster deployment of autonomous driving functions, driving the intelligent driving industry chain from vertical integration toward specialized division of labor. As a result, independent providers are seeing a sustained increase in both penetration rate and value share in the Mass-Production autonomous driving market.
Leading Chinese solution providers are driving the global development of Urban NOA: As technology maturity continues to improve, leading participants are leveraging partnerships with overseas OEMs and localized deployment capabilities to accelerate the introduction of mature Urban NOA solutions to global markets. In the future, highly transferable algorithm capabilities, engineering capabilities, and global delivery capabilities are expected to enable Chinese autonomous driving solution providers to evolve from domestic leaders into significant participants in the global Mass-Production autonomous driving industry, further advancing the global commercialization process of Urban NOA.
Competitive Landscape: a consolidating industry with two frontrunners
The Urban NOA field is exhibiting clear scale effects. As more vehicles achieve Mass-Production delivery, players can continuously accumulate real-world road data, enhancing model generalization capabilities and system safety; meanwhile, the accumulated Mass-Production experience across multiple vehicle models also helps establish standardized development and delivery systems, further reducing project implementation costs and improving delivery efficiency. The data, engineering, and customer resource barriers formed as a result are causing industry competition to gradually concentrate toward leading enterprises.
Currently, the independent solution provider market has formed a competitive landscape represented by Momenta and Huawei HI, with the two enterprises collectively accounting for more than 80% of market share, representing two distinct development pathways of the open platform model and the full-stack solution model, respectively.
Case study of two leading industry players in the Urban NOA: Huawei HI
Founded in 1987, Huawei is a global and China-leading provider of ICT infrastructure and smart devices. The company began developing its automotive business in 2009 and has since gradually built a smart automotive solutions portfolio covering intelligent driving, intelligent cockpits, connected vehicles, and cloud services. After years of investment, Huawei has formed five major product matrices centered around Qiankun ADS, Qiankun Vehicle Control Unit, Qiankun Vehicle Cloud, Automotive Photonics, and HarmonyOS Cockpit, possessing full-stack capabilities ranging from chips and computing platforms to sensors, software algorithms, and cloud services. Leveraging its long-accumulated ICT technology advantages and sustained R&D investment, Huawei has become an important participant in China’s in-house and third-party Urban NOA market and holds a significant position in the intelligent vehicle industry chain.
In terms of business models, Huawei simultaneously deploys two models – Huawei Inside (HI) and HIMA – to meet the needs of different OEMs. Under the HI model, the OEM leads product definition, while Huawei acts as an independent provider delivering full-stack smart automotive solutions, including intelligent driving, to automakers and enabling sales through OEM channels. This model currently covers around 7 OEM partners, including Avatr, BAIC, Voyah, Deepal, and others. In contrast, the HIMA model adopts a more integrated collaboration approach, where Huawei participates in product definition and key decision-making, deeply integrates autonomous driving solutions with vehicle development, and sells through Huawei’s own distribution channels. This has formed a brand portfolio including AITO, Luxeed, Stelato, Maextro, SAIC, etc. Together, these two models constitute the development framework of Huawei’s smart automotive business, enabling Huawei to expand its ecosystem coverage through the independent HI model while strengthening brand influence and product competitiveness via the in-house HIMA model, thereby delivering full-stack smart automotive solutions covering both hardware and software. From a sales structure perspective, the in-house model contributes the vast majority of Huawei’s smart automotive business sales volume and serves as the core vehicle for current commercial deployment.
Case study of two leading industry players in the Urban NOA: Momenta
Founded in 2016, Momenta is a leading autonomous driving independent solution provider globally and in China, focusing on the R&D and commercialization of autonomous driving technology, with a product portfolio covering two major business directions: Mass-Production autonomous driving and Scalable Robo. Although the Mass-Production autonomous driving solution and Scalable Robo operate independently, Momenta leverages its own strategy to achieve synergistic development between the two businesses. Mass-Production vehicles continuously contribute massive amounts of real-world road data and commercial revenue, driving algorithm iteration and technological advancement; meanwhile, the Scalable Robo further validates and expands autonomous driving capabilities, forming a virtuous cycle where data, technology, and commercialization mutually reinforce each other, accelerating the company's progress toward higher-level autonomous driving.
Momenta ranks in the industry's first tier in both technology and commercial deployment. At the commercialization level, the company ranks among the industry leaders in key metrics including the number of SOPs, number of SOPs and nominations and sales of vehicles in Urban NOA, demonstrating strong mass-production delivery and large-scale deployment capabilities. As of February 2026, Momenta held market shares of 57% and 65% in terms of SOPs and SOPs and nominations in Urban NOA. In addition, based on vehicle sales① from March 2025 to February 2026, Momenta accounted for approximately 65%② of the Urban NOA market as a clear industry leader. At the technical level, its Urban NOA system performs stably in typical urban scenarios, leading the industry average. Meanwhile, Momenta has pioneered the deployment of mass-production Robotaxi fleets in multiple cities globally and obtained Robotaxi commercial operation licenses. The Robovan business has also been steadily deployed, forming a "dual-wheel drive" development model covering both Mass-Production autonomous driving and Scalable Robo. It has become one of the important participants and representative enterprises in China's Urban NOA market.
The Future of Urban NOA: Toward Scalable L4
Although L4 autonomous driving is widely regarded as the industry's long-term development goal, from the perspective of industry evolution pathways, achieving large-scale commercialization still faces multiple challenges including safety validation, operational system construction, cost control, and regulatory framework improvement. In contrast, Urban NOA has achieved large-scale deployment in Mass-Production vehicles and is becoming an important bridge connecting current assisted driving and future fully autonomous driving.
Urban NOA is playing the role of a "training ground" for L4 autonomous driving. On one hand, large-scale Mass-Production vehicles continuously accumulate real-world road data, helping autonomous driving systems continuously improve their understanding and handling capabilities for complex scenarios; on the other hand, the long-term operation of Urban NOA through mass-production also drives enterprises to accumulate automotive-grade engineering capabilities, while helping users, OEMs, and regulatory agencies gradually build awareness and trust in autonomous driving technology. As the data closed loop continues to strengthen, technical capabilities continuously improve, and the industry ecosystem gradually matures, Urban NOA is expected to become the most realistic and commercially feasible pathway toward L4 autonomous driving.
Urban NOA provides real-world data on a scale. L4 autonomous driving needs to handle a large number of corner cases, but relying solely on dedicated fleets to accumulate data is constrained by fleet size, operational zones, and regulatory approvals. By entering a broad range of real-world user scenarios via mass-produced vehicles, Urban NOA covers more cities, more road types, more driving styles, and more behaviors of traffic participants, thereby building a much broader data foundation.
Urban NOA enhances model generalization capabilities. Complex intersections, unprotected turns, non-motorized vehicle weaving, avoiding parked vehicles, and construction detours – all scenarios that Urban NOA must handle – are precisely the problems L4 will need to solve in the future. The continuous operation of mass-produced Urban NOA in these scenarios helps the model learn the interaction patterns of real-world urban traffic, improving perception, prediction, planning, and control capabilities.
Urban NOA builds an automotive-grade engineering system. L4 is not merely an algorithm system; it also requires automotive-grade software, vehicle control interfaces, functional safety, OTA upgrade capabilities, sensor fusion, and a quality management system. The deployment of Urban NOA in mass-produced vehicles enables enterprises to accumulate experience in collaborative development with OEMs, vehicle validation, SOP delivery, and post-sales iteration. These engineering capabilities can be partially migrated to L4 platforms.
Urban NOA fosters user and regulatory awareness. User trust in autonomous driving is not built overnight. Urban NOA allows consumers to gradually become familiar with vehicle assisted driving capabilities, functional boundaries, and takeover mechanisms, while also enabling regulators to gain earlier exposure to how autonomous driving performs on real roads. This provides a trust foundation for future L3 and L4 responsibility allocation, insurance mechanisms, and road access.
Conclusion: Urban NOA is the starting point for commercialization, while L4 is the long-term endgame
The autonomous driving industry is moving from early technological imagination into a phase of pragmatic commercialization. Over years, the industry has debated whether autonomous driving will evolve incrementally through increasingly capable assisted driving systems toward Level 4 autonomy, or leapfrog directly to Level 4 through dedicated autonomous fleets — the former has been widely deployed but offers limited commercial value, while the latter represents the endgame but remains constrained by safety validation, regulatory approval, and cost, making large-scale deployment on open urban roads difficult in the short term. The strategic significance of Urban NOA lies precisely in bridging these two worlds: it leverages mass-production passenger vehicles to rapidly build an installation base and real-world user scenarios, while directly confronting core challenges close to L4 — such as complex urban roads, continuous driving tasks, and frequent human-vehicle interaction. Thus, Urban NOA is not only the most commercially visible form of autonomous driving today, but also an important precursor stage on the path to L4.
From an industry evolution perspective, Urban NOA addresses the question of “how to scale-up autonomous driving into mass-produced vehicles,” while L4 addresses “how to enable autonomous driving systems to truly replace human drivers in specific scenarios.” The former determines current market penetration, vehicle model-level competition, and the foundation for software-driven revenue; the latter determines the long-term technology ceiling, the boundaries of business models, and the restructuring of mobility and logistics services. More importantly, Urban NOA, through Mass-Production, accumulates real-world road data, user feedback, and complex-scenario experience, providing core support for the phased commercialization of L4.
Therefore, the long-term endgame of autonomous driving has not changed, but the path to that endgame is being reshaped. L4 still represents the long-term peak of the autonomous driving industry, yet Urban NOA is becoming the starting point where the industry truly moves toward scale, commercialization, and market repricing. In the future, enterprises with sustainable long-term competitiveness may not be those with the most test vehicles or the earliest feature releases, but those that can integrate mass-produced vehicle scale, real-world data, an automotive-grade engineering system, continuous software iteration, and a driverless technology roadmap.
①②：Sales and market share are calculated based on insurance registration sales volume
