Changan Proposes Automotive Engineering Approach to Robots, Aiming for Robots That Build Robots
On September 18, the embodied intelligence session of the 2026 TEDA Auto Forum's New Track Ecosystem special event was held, bringing automakers, robotics companies and researchers together around the emerging convergence of cars and robots.
Tan Huan, chief expert at Changan Automobile and general manager of Changan Tianshu Intelligent Robot Company, delivered the keynote address at the session.
Tan proposed that automobiles will evolve into embodied intelligence composite agents, and that the industry should use the automotive engineering system to empower the robotics industry while jointly building a broad mobility ecosystem.
He said intelligent vehicles of the future will develop into car-robot composite agents capable of self-learning, self-organization and self-evolution — a vision in which the car is no longer a single product but a platform that perceives, decides and acts alongside its users.
On data handling, Changan proposes a dual-drive model that combines knowledge and rules: knowledge and rules guide data collection, and the resulting data in turn feeds back to iterate and refine those rules. The approach is intended to keep data acquisition targeted and repeatable rather than open-ended.
The company also proposes transferring the automotive industry's SOP processes, layered algorithm decomposition and full-process quality control concepts into the robotics field. That thinking mirrors the methodology the auto sector has accumulated over decades in manufacturing and quality control.
After more than a century of development, automakers have built mature systems for standardized processes, quality management and supply chain coordination. Moved into robot manufacturing, these systems address the consistency problem that robots face when moving from prototype to mass production — an area where many humanoid robot developers still lack disciplined process engineering.
Changan's bet is that the lessons of large-scale car production transfer more directly to robot manufacturing than most robotics startups assume.