🐯虎嗅•Freshcollected in 21m
Humanoid Robots Enter Auto Factories
💡Humanoids are entering real factories—but costs must drop 60% before automakers can replace workers economically.
⚡ 30-Second TL;DR
What Changed
UBTECH reportedly delivered 1,079 full-size embodied-intelligence robots in 2025, with more than 80% entering industrial scenarios.
Why It Matters
Automotive manufacturing is becoming a critical proving ground for embodied AI, providing high-volume operational data and measurable return-on-investment benchmarks. The next adoption bottleneck is not demonstration capability but total cost of ownership, uptime, throughput, and integration with existing factory systems.
What To Do Next
Run a six-month pilot with Walker S or Galbot S1 on one repetitive station, logging uptime, cycle time, maintenance, and labor savings to calculate total ROI.
Who should care:Enterprise & Security Teams
Key Points
- •UBTECH reportedly delivered 1,079 full-size embodied-intelligence robots in 2025, with more than 80% entering industrial scenarios.
- •CATL, BYD, Geely, Dongfeng Liuzhou, and other automakers are testing humanoid robots in production environments.
- •Dongfeng Liuzhou estimates current robots cost ¥600,000–¥700,000 and need to reach about ¥250,000 to compete with human labor.
- •Current tests indicate a humanoid robot performs at roughly half the productivity of a human worker, while wheeled humanoids are more efficient than bipedal designs.
- •Deployment models include self-developed robots, third-party procurement, long-term factory training, and robot-as-a-service leasing.
🧠 Deep Insight
AI-generated analysis for this event.
🔑 Enhanced Key Takeaways
- •The Chinese Ministry of Industry and Information Technology (MIIT) issued guidelines in late 2025 establishing standardized safety and communication protocols for humanoid robots operating in shared human-machine workspaces.
- •Major battery manufacturers like CATL are integrating 'digital twin' synchronization, where humanoid robots are trained in a virtual environment before physical deployment to reduce factory downtime during the learning phase.
- •Recent supply chain data indicates that the cost reduction target of ¥250,000 is heavily dependent on the localization of high-torque actuators and harmonic drives, which currently account for 40% of the total BOM cost.
- •Beyond automotive, the 'Humanoid+AI' industrial model is expanding into hazardous chemical handling and high-voltage battery testing, where robots are equipped with specialized end-effectors for non-destructive inspection.
- •Local government subsidies in industrial hubs like Shenzhen and Hefei are currently covering up to 30% of the initial procurement costs for enterprises deploying domestic humanoid robots in pilot programs.
📊 Competitor Analysis▸ Show
| Feature | UBTECH (Walker S) | Tesla (Optimus Gen 3) | Figure AI (Figure 02) |
|---|---|---|---|
| Primary Focus | Industrial/Automotive | Mass Manufacturing | General Purpose/Logistics |
| Estimated Cost | ¥600k - ¥700k | ~$20k - $30k (Target) | High (Enterprise) |
| Key Benchmark | High integration with MES | FSD-derived neural nets | OpenAI-powered reasoning |
| Deployment | Active in Chinese OEMs | Internal Tesla pilot | Pilot in BMW/Logistics |
🛠️ Technical Deep Dive
- Integration of Large Vision-Language Models (LVLMs) allows robots to interpret natural language instructions for task re-sequencing without manual code updates.
- Utilization of multi-modal sensor fusion combining LiDAR, depth cameras, and tactile feedback sensors in fingertips for precise handling of delicate battery components.
- Implementation of Whole-Body Control (WBC) algorithms that enable robots to maintain balance while performing asymmetric tasks like tightening bolts or lifting heavy parts.
- Adoption of modular actuator designs that allow for rapid field replacement, reducing mean time to repair (MTTR) in high-uptime factory environments.
🔮 Future ImplicationsAI analysis grounded in cited sources
Humanoid robot adoption will reach a 15% penetration rate in Chinese EV assembly lines by 2028.
The current trajectory of cost reduction and the integration of standardized industrial communication protocols will likely overcome the initial economic barriers within the next 24 months.
The 'Robot-as-a-Service' (RaaS) model will become the dominant procurement method for SMEs.
High upfront capital expenditure remains the primary barrier, making leasing models more attractive than direct ownership for smaller automotive suppliers.
⏳ Timeline
2023-12
UBTECH Robotics completes IPO on the Hong Kong Stock Exchange.
2024-02
UBTECH announces the Walker S humanoid robot for industrial applications.
2024-05
Strategic partnership established between UBTECH and Dongfeng Liuzhou for factory pilot testing.
2025-01
UBTECH reports delivery of over 1,000 embodied-intelligence robots for industrial use.
2025-11
MIIT releases national standards for humanoid robot industrial safety.
📰
Weekly AI Recap
Read this week's curated digest of top AI events →
👉Related Updates
AI-curated news aggregator. All content rights belong to original publishers.
Original source: 虎嗅 ↗