The Physical AI startup, backed by Yunqi Capital and SenseTime, is targeting logistics “hand-eye coordination” tasks and plans to prioritize Japan and South Korea in overseas expansion.
Li Qiang, former CTO of Cainiao Group and CTO of Alibaba International Digital Commerce Group, has founded Hangzhou Quantum Dynamics Intelligent Co., Ltd., a Physical AI company focused on linking AI with real-world operations. The startup has raised a seed round of more than 100 million yuan, or about $14.8 million, from Yunqi Capital and SenseTime, with funds earmarked for core technology research and development, hiring and global expansion. Duowei Capital helped incubate the project and assemble the team. The company is starting with difficult logistics tasks that require “hand-eye coordination,” such as weighing parcels, applying labels, shelf restocking, picking and packing, an area Li said older automation systems could not solve at scale. He said previous technology addressed only “twenty to thirty percent” of the problem set and that new Physical AI technologies in 2025 made a broader solution feasible. Quantum Dynamics is building what it describes as an embodied native world model, combining vision, touch and force feedback to predict environmental change, generate actions and assess state value. The company’s approach also centers on a data flywheel (self-reinforcing improvement loop) built from robots operating in real-world settings, with edge-side deployment (computing on the device) used to capture and route high-value training data. Li argued that, in Physical AI, data contribution will exceed 50% and that scalable deployment depends not just on single-robot capability but on swarm intelligence (multi-robot coordination). Commercially, the startup plans to begin in B2C logistics warehouses and distribution centers, then expand into semi-open retail outlets and eventually last-mile delivery. Li said the company is in talks with several leading logistics enterprises, is close to signing a strategic cooperation agreement with a major domestic integrated logistics company for multi-warehouse picking pilots, and has secured adoption of its picking solution by the world’s largest health supplement e-commerce platform for overseas warehouses, with first deployments expected within the year. He aims to reach PMF (product-market fit) in one to two scenarios by the end of 2026 and have more than 1,000 robots working daily in the physical world by 2027.