DeepRoute.ai Advanced Driver Assistance System breakthrough: over 300k vehicles deployed. 2026 target: 1 million City NOA fleet.

Chinese self-driving startup DeepRoute.ai announced on 4/25 to the public that its advanced driver assistance system has been deployed in more than 300k vehicles within mainland China. Reuters, citing DeepRoute.ai’s data, reported that over the past year, the system has helped avoid more than 180k potential collision incidents. The company also publicly released its 2026 targets: delivering more than 1 million City NOA (automated navigation driver assistance) vehicles; raising high-frequency utilization to 50% or above, serving as key indicators of its push toward large-scale commercialization of Robotaxis.

The speed of penetration from driver assistance to City NOA

The cumulative deployment of 300k vehicles in the Chinese market has structural significance. China’s advanced driver assistance driving competitions have at least narrowed down to several leading players. DeepRoute.ai, along with Huawei and Momenta, are the three vendors with the largest scale and breadth of coverage at the OEM level. This data shows the pace at which it has entered mainstream mass-production vehicle models—expanding from single partnership vehicle models in 2024 to multiple flagship models across multiple automakers. City NOA is the advanced driver assistance function module with the greatest market penetration potential in China—it corresponds to the “no takeover for the entire commute” scenario in everyday commuting. Compared with Highway NOA, it has higher usage frequency and user stickiness.

Dual goals for 2026: 1 million deliveries + 50% utilization rate

For DeepRoute.ai, the two 2026 figures it has disclosed must be achieved simultaneously for them to be meaningful. 1 million vehicles represents breadth on the supply side, while a 50% high-frequency utilization rate represents depth on the demand side—meaning users truly turn on and rely on this feature in real-world driving, rather than the “car has the function but very rarely uses it” situation. The commercial viability of Robotaxis needs to be built on training and data foundations based on the premise that “human drivers have become accustomed to letting go.” Therefore, these two numbers together point to the feasible window for large-scale Robotaxi deployments in 2027–2028.

Signals for the self-driving landscape in China and the U.S.

DeepRoute.ai’s progress timing aligns with the synchronized push of self-driving infrastructure between China and the U.S., following Tesla’s Cybercab starting mass production earlier this week. On one side is the U.S. automaker’s vertical integration (Tesla manufactures vehicles, builds its own models, and operates its own fleet). On the other side is China’s software vendor deployment model (DeepRoute.ai partners with multiple automakers, with each deploying on existing vehicle models). At the scaling stage, these two architectures will diverge in cost structures and regulatory pathways—one of the key dimensions for observing the self-driving industry.

For readers in the Asia-Pacific market, DeepRoute.ai’s data also implies that advanced driver assistance in mainland China has already entered a “normalized usage” stage rather than pilot experiments. For observers of customer composition in Taiwan and Southeast Asia’s auto supply chains and component-factory clients, this will provide a more concrete timeline anchor.

This article, “DeepRoute.ai’s advanced driver assistance system breakthrough: 300k-vehicle deployments—2026 target 1 million City NOA fleet,” first appeared on Lian News ABMedia.