On September 16-17, the Convergent Intelligence Industry Development Conference (2025) was held in Wuhan, themed "Collaborative Integration of Innovation Chain and Industry Chain to Promote Convergent Intelligence Industry Development." Zhang Hanwen, Ecosystem General Manager of IMOTIONTECH's Robotics Division, introduced that the company was founded in 2016 and is China's first listed autonomous driving solution enterprise. Headquartered in Suzhou, the company has algorithm and R&D centers in Beijing, Shanghai, Guangzhou, Wuhan, as well as Germany and Malaysia.

He mentioned that the company's core advantage lies in providing fully self-developed integrated software and hardware solutions. Unlike pure software or hardware manufacturers, IMOTIONTECH develops everything in-house from perception and decision-making to control, covering integrated parking and driving domain control hardware and core algorithms. "Because of this, we have gained recognition from almost all mainstream OEMs except for new energy vehicle startups. We recently reached a strategic partnership with a leading new energy vehicle company, and related mass production projects will be implemented gradually."

Why enter the robotics industry? Zhang Hanwen explained that the core reason is the high overlap between smart vehicles and embodied intelligence in both technology and supply chains.

First, at the controller level, he noted that this year the company's intelligent driving domain controller deliveries will exceed 650,000 units, covering multiple chip platforms including NVIDIA, TI, Renesas, and Horizon Robotics. The company has particularly deep cooperation with Horizon Robotics, achieving mass production deployment from the J3 to J6 series. He believes that the reliability, real-time capabilities, and redundancy of automotive-grade controllers perfectly match the "large-small brain collaboration" requirements of robots, especially embodied intelligence systems.

Second, at the algorithm level, the company's engineering optimization experience on low-computation platforms can be reused in robotic environments. With core components and algorithms in place, physical verification is still needed. Earlier this year, IMOTIONTECH acquired a joint development company in Suzhou called "Little Craftsman" and established AIMOSTAR Robotics based on this acquisition, formally launching robot R&D and commercialization through a business division model. "We developed an industrial-grade welding robotic arm to primarily verify its load capacity, repeat positioning accuracy, and multimodal perception fusion capabilities. Test results have been favorable, initially proving our capability in physical development and system integration."

Regarding controllers, Zhang Hanwen stated that the robotics field currently generally adopts NVIDIA solutions, which, while having mature ecosystems and powerful computing capabilities, also present issues of high costs and supply chain risks. Therefore, in June this year, the company partnered with Horizon Robotics and successfully created the world's first domestic medium-to-high computing power robot controller based on Horizon's "Sweet Potato" robot development platform.

Based on the above, IMOTIONTECH developed a six-axis collaborative robotic arm for automatic charging scenarios. This robotic arm uses a monocular vision pose detection algorithm running entirely on domestic chip edge computing. "At last week's Munich Motor Show in Germany, we conducted our first public demonstration, completing one plug insertion action in just 19.5 seconds, outperforming similar industry solutions," he said.

Zhang Hanwen also mentioned that from a broader technological wave perspective, from Fairchild Semiconductor to today, each decade of technological iteration has brought exponential growth. "The emergence of ChatGPT in 2022 accelerated the explosion of AI application ecosystems. We believe that just as large language models gave birth to numerous agents and applications, robots will become the physical carriers of the next generation of AI intelligent agents."

Finally, he emphasized that only by combining advanced algorithms with powerful physical bodies can true embodied intelligence be achieved. "As an intelligent driving company, we will actively embrace this trend and work with the industry to promote the scaling and intelligent development of robots."