Xiaomi Auto L3 autonomous driving Lead Leaves for AI

Wang Naiyan Leaves Xiaomi Auto for Physical AI Venture

Xiaomi Auto L3 autonomous driving technology lead Wang Naiyan has reportedly left Xiaomi to launch a physical AI venture, marking another move by a senior Chinese automotive technology specialist into robotics and embodied artificial intelligence. Wang has recently met with investors while exploring opportunities in physical AI, according to 36Kr. Separately, 21jingji reported, citing multiple independent sources, that Wang has launched an embodied AI venture focused on robot brains and is seeking funding. The venture reportedly includes several former colleagues from TuSimple, now renamed CreateAI, with more than 10 core researchers and engineers, according to an investor cited by 21jingji.

Wang's Background in Autonomous Driving Technology

Wang holds a doctorate from the Hong Kong University of Science and Technology and spent nine years working in TuSimple's China operations. He became chief technology officer for TuSimple's China region in 2019, according to 36Kr. Wang joined Xiaomi Auto in May 2024, where he was responsible for autonomous driving technology development and L3 projects. He reported to Ye Hangjun, general manager of Xiaomi's autonomous driving and robotics division. His departure follows the exit of Chen Long, meaning two of the four core members of Xiaomi's autonomous driving team announced last year have now left to establish startups, according to 36Kr.

Autonomous Driving Experience Moves Into Embodied AI

Wang's departure is part of a broader migration of autonomous driving specialists from China's automotive sector into robotics and physical AI. Former executives and senior technical leaders from Xpeng, Li Auto, and Nio Inc have either joined robotics companies or started independent ventures. The movement reflects an effort to transfer capabilities developed through autonomous driving, including perception, planning, and reinforcement learning, into systems designed to operate in a wider range of physical environments. For automakers, the trend also creates additional pressure to retain experienced AI researchers and engineers whose expertise can potentially be applied across multiple emerging technology fields.

Wang's Xiaomi Work Included Nürburgring L3 Testing

One of Wang's notable projects at Xiaomi involved an autonomous driving demonstration at the Nürburgring Nordschleife in June. At the time, Wang said that exploring the limits of vehicle performance could support L3 development by providing a better understanding of the physical boundaries that autonomous driving systems must operate within. His work at Xiaomi therefore connected vehicle performance testing with the development of higher-level autonomous driving capabilities. The experience could also provide a technical foundation for his new direction, although the extent to which automotive autonomy expertise can translate into robotics remains uncertain.

Physical AI Venture Targets More Generalized Robot Models

Wang's new venture is expected to focus on embodied AI foundation models designed to provide stronger generalization capabilities, according to 21jingji, which cited people close to the project. The objective is to enable systems to adapt to environments and tasks that were not specifically covered by their training data. The project reportedly plans to introduce reinforcement learning during action-model pretraining, allowing models to learn behavioral strategies through extensive trial and error in simulations. The venture is expected to concentrate initially on developing foundational capabilities before applying those models to specific physical-world scenarios.

Former Automotive AI Leaders Are Joining Robotics

Several other moves illustrate the same shift across China's technology and automotive sectors. Former Xpeng autonomous driving chief Li Liyun and former vice president Chen Yonghai have joined EngineAI Robotics. Former Li Auto chief AI scientist Chen Wei founded Xieyue Intelligent, which focuses on consumer embodied AI products for home use. Li Auto's former autonomous driving chief Lang Xianpeng also left and co-founded Kunlunxing Robotics with former Huawei executive Ren Geng. These developments show how autonomous driving leadership experience is increasingly becoming relevant to companies developing robots and other embodied AI systems.

Nio Takes a Different Approach to Physical AI

Nio Inc has adopted an arrangement intended to retain a key autonomous driving executive while supporting an external physical AI venture. Nio founder, chairman, and CEO William Li confirmed during the company's September 1 earnings call that autonomous driving chief Ren Shaoqing would launch a physical AI venture with Nio acting as a strategic shareholder. Ren will remain head of Nio's autonomous driving division and continue to oversee its technical direction and long-term strategy. Li said the arrangement allows Nio to maintain strategic investment in physical AI while concentrating on its core business without affecting its profit and loss.

China's Automotive AI Talent Faces New Opportunities

The movement of technical leaders from China's automotive industry into physical AI reflects the growing overlap between autonomous vehicles and robotics. Both fields require advanced capabilities in perception, planning, decision-making, simulation, and reinforcement learning, but transferring those skills is not straightforward. Industry sources cited by 21jingji noted that robots face more complex motion-control requirements, while data collection and simulation training can also be more difficult. The emergence of new ventures could therefore create opportunities for experienced automotive AI teams, while forcing automakers to compete more directly for researchers capable of developing advanced physical-world intelligence systems.

Industry Impact & Outlook

The movement of autonomous driving specialists into physical AI could reshape how automotive companies compete for advanced AI talent, particularly as the technical boundaries between autonomous vehicles and robotics become less distinct. Xiaomi Auto now faces the loss of another core autonomous driving team member, while Nio's approach demonstrates an alternative model that combines executive retention with strategic investment in an external venture. The immediate significance is less about the number of startups than the potential redistribution of specialized expertise across industries. The success of these ventures will depend on whether autonomous driving methods can be adapted effectively to robotics' more demanding motion-control, data, and simulation requirements.

Frequently Asked Questions

Why did Wang Naiyan leave Xiaomi Auto?
Wang Naiyan reportedly left Xiaomi Auto to pursue a physical AI venture focused on embodied AI foundation models and robot brains. He has reportedly been meeting investors and seeking funding for the new company, which includes former TuSimple colleagues and more than 10 core researchers and engineers. The venture plans to use reinforcement learning during action-model pretraining and initially develop foundational capabilities before applying them to specific scenarios. His departure follows another exit from Xiaomi's four-member core autonomous driving team.