AI Agents From RMB 98,000 a Unit to a 10,000-Unit Factory: Sichuan Embodied Tech Raises Over RMB 100 Million for Its Emotion-Reading Humanoid News From 3 Councils to 15: DeepMind PlanAssist Enters a 12-Month Beta After Cutting UK Planning Assessment Time by 58 Percent Security OpenAI Will Train on Daybreak Blue Cybersecurity Session Data From November 29, Reversing a No-Training Guarantee News 16 Months and 2.87 Million RMB Later: Top Chinese Convenience Chain Hongqi Killed Its Robot Store Clerk Project AI Agents Westlake University Opens a World Model Research Center as Its Startup Awomo Claims Four Robotics Benchmark Wins Apple Intelligence Apple Quietly Licensed the IP of Agentic Data Engineering Startup Ascend.io and Hired Some of Its Team, EU Filing Shows Productivity 90% of Its Code Is Already Written by AI: Proton Cancels Its Controversial Employee Re-Interviews After Backlash Claude Two Sentences Now Stop Claude Code From Rebuilding GTA or Pokemon: Anthropic Bans Recreations of Existing Games AI Agents From RMB 98,000 a Unit to a 10,000-Unit Factory: Sichuan Embodied Tech Raises Over RMB 100 Million for Its Emotion-Reading Humanoid News From 3 Councils to 15: DeepMind PlanAssist Enters a 12-Month Beta After Cutting UK Planning Assessment Time by 58 Percent Security OpenAI Will Train on Daybreak Blue Cybersecurity Session Data From November 29, Reversing a No-Training Guarantee News 16 Months and 2.87 Million RMB Later: Top Chinese Convenience Chain Hongqi Killed Its Robot Store Clerk Project AI Agents Westlake University Opens a World Model Research Center as Its Startup Awomo Claims Four Robotics Benchmark Wins Apple Intelligence Apple Quietly Licensed the IP of Agentic Data Engineering Startup Ascend.io and Hired Some of Its Team, EU Filing Shows Productivity 90% of Its Code Is Already Written by AI: Proton Cancels Its Controversial Employee Re-Interviews After Backlash Claude Two Sentences Now Stop Claude Code From Rebuilding GTA or Pokemon: Anthropic Bans Recreations of Existing Games

From RMB 98,000 a Unit to a 10,000-Unit Factory: Sichuan Embodied Tech Raises Over RMB 100 Million for Its Emotion-Reading Humanoid

From RMB 98,000 a Unit to a 10,000-Unit Factory: Sichuan Embodied Tech Raises Over RMB 100 Million for Its Emotion-Reading Humanoid

Sichuan Embodied Technology, a Chengdu robotics startup founded in 2024 by a team from the University of Electronic Science and Technology of China, has raised more than RMB 100 million, about 14 million dollars, in a Pre-A round, according to West China Metropolis Daily. The money goes toward upgrading Aiqiu, a humanoid with a 3D projected face that the company says recognizes 15 emotions, and building a 10,000-unit production base.

Sichuan Embodied Technology, a Chengdu humanoid robotics startup, has completed a Pre-A funding round of more than 100 million yuan, roughly 14 million dollars, the West China Metropolis Daily reported on November 2. The company says the capital will fund iteration of its flagship emotional interaction robot Aiqiu, construction of a production base with a capacity of 10,000 units a year, and the build-out of a national sales network. A new funding round is already underway, according to co-founder and CEO Feng Zhenyu.

Founded in September 2024 by a core team from the University of Electronic Science and Technology of China, the company pursues a full-stack approach covering the robot body, core algorithms and key components. Alongside the round it has obtained project support from Sichuan Provincial Pilot-Scale R&D, a state-backed entity, creating what it describes as coordinated support from provincial, city and county-level state capital.

Aiqiu, the product the round is named for, made its debut at the 2026 World Robot Conference in August, where the company says it drew large crowds and multiple intent orders from home and abroad. The robot uses 3D projection for its face and, by the company's claim, can accurately recognize 15 categories of human emotion. It carries an official starting price of 98,000 yuan, about 13,800 dollars, which places it well below the research-grade humanoids that dominate most conference stages.

The company says all three of its product lines are now in batch delivery: Aiqiu for emotional interaction services, Tianxun for intelligent inspection, and Tianxingzhe for robot training and research testing. Its development loop runs through a provincial humanoid robot training ground built jointly with the Tianfu Jiangxi Laboratory, combining real-machine data collection, data governance, model training, evaluation and deployment with data flowing back for iteration.

The funding lands in a market scaling fast on paper. Figures cited by the Ministry of Industry and Information Technology put China's humanoid robot output above 40,000 units in the first half of 2026, with the full year expected to exceed 100,000. Chengdu is trying to build a local ecosystem around that curve; the same report notes a museum guide robot made by another local firm, Tianjiang Xinhua, that has served more than 100,000 visitors at the Chengdu Natural History Museum.

The usual caveats apply. The story rests on a single local newspaper report and company statements: the 15-emotion recognition claim, the intent orders and the delivery status of all three product lines are company figures that have not been independently verified, and no investors in the round were named. Emotional companion robotics is also an unproven consumer category, and a 98,000-yuan price tag is a bet that families will pay research-adjacent money for what is, functionally, a very expressive display on a mobile body.

What the round does confirm is where Chinese humanoid capital is heading. Investors in the sector have been blunt that dancing demos no longer raise money; the question asked now is whether a robot holds a real job. Sichuan Embodied is betting the answer includes jobs where the output is feeling rather than force: greeting, accompanying and entertaining. Whether that counts as a real scenario is exactly what the next round will have to prove.

Comments (0)

Log in to join the discussion

Log In

No comments yet