This funding is a significant milestone for the five-year-old company, which is positioning itself to compete with established players like Tesla in the race toward autonomous vehicle solutions, CEO Maxwell Zhou announced.
Like Tesla, Deeproute’s approach to self-driving technology is data-driven, relying on self-learning artificial intelligence instead of traditional, rule-based systems. According to Zhou, China’s vast pool of complex traffic data gives Deeproute a distinct advantage in developing autonomous systems specifically suited for the unique demands of Chinese roads.
“We’ve been training our system using Chinese traffic data for a long time,” he said, pointing out that Tesla is still awaiting regulatory approval to introduce its Full Self-Driving (FSD) system to China next year.
Zhou’s announcement comes alongside a recent investment by Great Wall Motor, with a 7% stake in Deeproute through its affiliate, Baoding Xunqi Technology Co., Ltd., according to official filings in Shenzhen. Alibaba remains Deeproute’s largest shareholder with a 17.2% stake.
Zhou highlighted Deeproute’s upcoming “end-to-end” AI technology, which can handle complex city driving environments with minimal human intervention—an essential feature as Chinese cities present some of the world’s most challenging traffic scenarios. Deeproute’s current system, Deeproute IO 2.0, is already implemented in 20,000 vehicles, including Geely’s Smart #5 and Great Wall’s Blue Mountain SUV. By next year, the company expects to roll out its AI-enabled systems to over 10 new models, bringing advanced assisted driving capabilities to the mainstream market.
The potential of NVIDIA’s next-generation DRIVE Thor chips also plays a key role in Deeproute’s strategy. Zhou believes the extra computing power provided by these chips will allow Deeproute’s AI to process larger amounts of raw sensor data, thereby improving the quality and responsiveness of its autonomous driving technology.
Deeproute’s data-focused approach sets it apart from conventional self-driving tech that relies heavily on high-definition mapping and lidar sensors. By forgoing these high-cost components, Deeproute’s solution is expected to be highly scalable and cost-effective, with a full hardware and software package projected to cost between US$1,000 and US$2,000 per vehicle.
“We’re not just chasing the technology but are focusing on creating an AI that understands and adapts to real-world driving complexities,” Zhou said, highlighting the scalability of the system for both Chinese and overseas markets while complying with data security regulations.
Zhou envisions that by 2024, autonomous driving technology—especially in the form of ADAS (advanced driver assistance systems)—will become a “killer feature” in electric vehicles across China.
As Deeproute builds out its partnerships and leverages NVIDIA’s DRIVE Thor chips, Zhou anticipates that the company’s offerings will bring advanced autonomous capabilities closer to the everyday driver, positioning Deeproute as a leader in making autonomous driving accessible, adaptable, and reliable in the world’s largest auto market.
