Industrial robotics gains a funding jolt from two Chinese startups

Industrial robotics gains a funding jolt from two Chinese startups. Inverse Matrix Technology and SUNRISING AI have both announced multi-million‑dollar rounds in short succession, underscoring a China‑driven push into embodied AI for factory floors. Inverse Matrix says it raised over $10 million in its first funding round, backed by Hillhouse Ventures and Yanyuan Ventures, with Gaohu Capital serving as exclusive financial advisor. The team, led by Ji Jiaming—PhD candidate at Peking University’s Institute for AI—and Chen Boyuan, an undergraduate from Yuanpei College, has grown to more than 30 researchers. The company’s aim is to build world models (世界模型) that understand physical laws and generate predictions aligned with real-world dynamics, blending reinforcement learning with world models to tackle embodied AI, industrial simulation, and scientific computing. The plan: debut a flagship model in 2026.

SUNRISING AI tells a complementary story: more than $14 million (roughly RMB 100 million) raised within six months of its founding, with a round led by IDG Capital and Oriental Fortune Capital and participation from EFORT Intelligent Equipment, 01VC, DaTai Capital, and L2F Ventures. Founded in April 2025 by Zhang Tao, a former Alibaba AutoNavi technical director, and Li Shengbo, a professor at Tsinghua University, SUNRISING AI focuses on wheeled industrial robots for automotive final assembly—an arena many say sits at the heart of China’s $100‑billion‑plus automotive manufacturing opportunity. The company has developed a high‑smoothness neural network architecture for industrial manipulation, leaning on reinforcement learning rather than imitation learning to improve precision and success rates. To address data shortages, SUNRISING AI leans on simulated data and has built its GOPS platform to modularize model development for scalable deployment. They have already partnered with multiple automakers and completed initial proof‑of‑concept deployments, with plans to reach at least 10 automotive manufacturers and deploy more than 1,000 robots within three years.

The two rounds illuminate a broader pattern: private venture funds are funneling capital into China’s next‑generation factory AI. These are not state‑owned pushes but capital markets signaling a belief that software‑defined intelligence can materially reduce the cost per robot, shorten cycle times, and raise yields on mature assembly lines. The investments come as manufacturers continue to seek automation that can operate in complex, semi‑structured environments—precisely where world models and embodied AI promise a edge over traditional control methods.

Two practical implications stand out for global manufacturers considering China as a sourcing or supply‑chain partner. First, the data challenge that hardware‑centred automation has long faced is being addressed with a dual strategy: building robust world models (inverse dynamics and physical reasoning) on the software side, and coupling that with real‑world experimentation on the shop floor. Inverse Matrix’s emphasis on world models and SUNRISING AI’s emphasis on sim‑driven data and modular deployment both point to a future where factory automation becomes more theory‑driven and less hand‑tuned. This matters for OEMs that want predictable automation performance across multiple sites and line configurations.

Second, the channel dynamics are shifting. Private investors are backing teams with deep academic roots and OEM‑level partnerships, signaling a go‑to‑market path built around joint development with automotive manufacturers rather than standalone products. SUNRISING AI’s roadmap—10 automakers and 1,000 robots in three years—illustrates the scale and the cooperative model needed to realize it. For global manufacturers, these moves translate into potential access to domestic Chinese supplier ecosystems that can deliver highly customized automation stacks, while also imposing a risk of intensified competition as more Chinese firms reach production‑grade capabilities.

Two key caveats remain. Even as simulation‑first strategies mitigate real‑world data gaps, the sim‑to‑real transfer remains a delicate boundary that can throttle early deployments. And while these rounds signal confidence, actual factory performance depends on durable OEM partnerships, supply chain resilience for robot components, and the ability to scale software upgrades across thousands of units.

Still, the direction is clear: China’s embodied AI labors on, turning research into factory reality at a pace that could redefine global auto and general manufacturing benchmarks.

Sources

Inverse Matrix Technology Raises Over $10 Million to Advance World Models

SUNRISING AI Raises Over $14 Million Within Six Months of Founding