Huawei's Jin Yuzhi reiterated that L3 is the critical bridge from assisted driving to full autonomy and cannot be skipped. Huawei has invested 18 billion yuan in Qiankun ADS, with system installations exceeding 1.7 million vehicles and cumulative assisted-driving mileage surpassing 10 billion kilometers.
The Core Significance of L3: A Technical Watershed in Liability Transfer
In the evolution of autonomous driving technology, Huawei is attempting to define an "unskippable" path. Jin Yuzhi, senior vice president at Huawei and CEO of Yinwang Intelligent, recently stated clearly that L3 is the critical bridge from assisted driving to full autonomy—a stage that cannot be bypassed in terms of either technical or industrial logic.
Jin's core argument is that the transition from L2 to L3 is not merely a technical upgrade, but a fundamental shift in liability—accident responsibility transfers from the driver to the vehicle manufacturer or autonomous system provider. This shift requires not only breakthrough technology, but also supporting legal frameworks and time for users to adapt. Therefore, a transitional phase is essential for driving the industry forward.
He further noted that L3 and L4 are not mutually exclusive, but distinct milestones on the same trajectory toward full autonomy. To achieve an L4 safety level more than ten times better than human driving, the only path is accumulating real-world road data and helping users gradually adapt through the L3 stage.
Huawei Qiankun's Technical Foundation: Data and Scale as Dual Drivers
Huawei has built its competitive moat for the Qiankun intelligent driving system on two key numbers:
- Fleet Scale: As of April 2026, Qiankun ADS has been installed in 1.7 million vehicles, with projections to exceed 3 million by year-end, covering 28 vehicle platforms including Avatr, Deepal, Voyah, and Harmony Intelligent Mobility
- Data Mileage: Cumulative assisted-driving mileage has surpassed 10 billion kilometers, with over 280 million automated parking maneuvers and 4.79 million potential collisions avoided by the active safety system
In terms of safety performance, Huawei's "Safe Mobility Report" shows that vehicles equipped with Qiankun ADS achieve an average of 5.17 million kilometers between severe collisions (requiring airbag or seatbelt pretensioner activation) in human-driven mode—2.87 times the national average of 1.8 million kilometers. In assisted-driving mode, this extends to 7.57 million kilometers, representing a 4.2-fold safety improvement.
| Metric | National Average | Qiankun ADS Human Mode | Qiankun ADS Assisted Mode |
|---|---|---|---|
| Severe Collision Interval | 1.8M km | 5.17M km | 7.57M km |
| Safety Multiplier | 1.0x | 2.87x | 4.2x |
| Data Source | China Road Statistics | Huawei Qiankun System | Huawei Qiankun System |
Industry Roadmap: Three Waves and the 2027 Inflection Point
Huawei divides autonomous driving development into "three waves":
- First Wave (2015–2022): Electrification, focused on reducing travel costs
- Second Wave (2022–2027): Intelligence, targeting safety and driver workload reduction
- Third Wave (2027+): Automation, redefining safety standards and enabling large-scale deployment of robotaxis and logistics fleets
The timeline is already quite specific:
- 2026: Highway L3 road testing underway in seven cities—Beijing, Shanghai, Guangzhou, Shenzhen, Hefei, Wuhan, and Chongqing—with pilot programs launching this year
- 2027: Large-scale commercial deployment of highway L3, with urban L4 pilots starting simultaneously
- Long-term: L3 achieves mainstream adoption while L4 enters early commercialization
Notably, Huawei's investment in the Qiankun ecosystem reached 18 billion yuan (approximately $2.6 billion) in 2026, with roughly 10 billion yuan earmarked for computing infrastructure to train next-generation autonomous systems. Jin has declared 2026 "Year One of Global Autonomous Driving," noting that both China and the United States are pushing toward L3 commercialization in parallel.
Industry Competition: From Route Debates to Ecosystem Games
Huawei is not the only player betting on L3. China's Ministry of Industry and Information Technology approved the first two batches of L3 models for road pilots in 2025, with 2026 seeing an orderly rollout of additional approvals. OEMs including GAC, Changan, Voyah, IM Motors, XPeng, and FAW-Volkswagen have all disclosed progress in L3 deployment.
However, differing viewpoints exist within the industry. Zhuoyu Technology CEO Shen Shaojie argues that in the large-model era, L2, L3, and L4 can evolve from a shared technical foundation, and advancing directly to L4 would help clarify liability boundaries. Huawei's gradual approach, however, has found policy-level resonance—joint guidelines from MIIT and the Ministry of Justice explicitly state that manufacturers bear primary responsibility for L3 and higher autonomous driving incidents, providing crucial legal support for L3 technology.
Jin also revealed that Huawei is collaborating with three insurance companies to explore insurance innovations related to intelligent driving. He believes competition among automakers has extended into downstream areas like insurance and after-sales services, but safety remains the bottom line—personal injuries cannot be remedied merely by repairing vehicles.
From the perspective of overseas buyers and dealers, Huawei's technical accumulation and scale advantages in intelligent driving are becoming a key differentiator for Chinese smart vehicle exports. For buyers in Central Asian and Russian markets who prioritize intelligent driving features, understanding the maturity and deployment timeline of China's L3 technology will be a critical variable in procurement decisions. For more intelligent driving technology updates, visit EX1000.COM.
