Shengshi Weisheng Completes Hundreds of Millions in B-Round Financing with Self-Developed Welding Embodied Brain Model
3Srobotics (Shengshi Weisheng) secured hundreds of millions in RMB Series B funding led by Shanghai Semiconductor Industry Investment and Jinqiao Fund to scale its industrial embodied AI solutions. The company utilizes a proprietary "autonomous brain" (multimodal large model for welding process decision-making) and "autonomous small brain" (real-time motion control system) architecture, enabling robots to adapt to non-standard complex scenarios without pre-programming. By targeting the critical
Analysis
TL;DR
- 3Srobotics (Shengshi Weisheng) secured hundreds of millions in RMB Series B funding led by Shanghai Semiconductor Industry Investment and Jinqiao Fund to scale its industrial embodied AI solutions.
- The company utilizes a proprietary "autonomous brain" (multimodal large model for welding process decision-making) and "autonomous small brain" (real-time motion control system) architecture, enabling robots to adapt to non-standard complex scenarios without pre-programming.
- By targeting the critical labor shortage in welding, their solution offers a plug-and-play capability that replaces 1.5-2 welders per unit with a 1-1.5 year ROI, addressing high non-standardization rates in industries like shipbuilding and steel structures.
- The business model combines direct sales for benchmark cases with channel distribution and an emerging "robot-as-a-service" pricing structure based on welded meterage, aiming for hundreds of millions in revenue by 2026.
- Strategic expansion plans involve leveraging the welding core competency to develop a full ecosystem of industrial intelligent robots for grinding, cutting, assembly, and handling, moving from single-scenario products to comprehensive manufacturing workflows.
Why It Matters
This development highlights the maturation of embodied AI from theoretical research to tangible industrial application, specifically demonstrating how multimodal models can solve long-standing problems in non-standardized manufacturing environments. For AI practitioners and investors, it underscores the value of vertical integration—combining hardware, software, and domain-specific data—to create defensible moats in sectors where generic automation fails. Furthermore, it signals a shift in the industrial robotics market towards "smart" units that offer rapid deployment and continuous learning capabilities, addressing urgent demographic challenges in skilled labor availability.
Technical Details
- Autonomous Brain Architecture: A multimodal large model trained on tens of millions of real-world industrial welding data points, including material properties, joint geometries, and environmental conditions. It enables real-time 3D reconstruction, spatial understanding, and dynamic generation of welding parameters (current, voltage, speed, swing amplitude) without relying on fixed programming.
- Autonomous Small Brain Control: A proprietary real-time motion control system developed through subsidiary Hangke Modern, ensuring low-latency execution of the brain's decisions. It provides precise trajectory correction and dynamic error compensation for complex shapes, deformations, and narrow spaces, overcoming the limitations of generic off-the-shelf controllers.
- Data-Driven Closed Loop: The system employs a self-learning mechanism where operational data continuously feeds back into the model, creating a positive flywheel effect ("more usage leads to more data, which leads to smarter models"). This allows the robot to improve its welding quality and efficiency over time.
- Hardware Integration: The solution integrates high-precision 3D vision systems with custom-built robotic arms, supporting various form factors including fixed and tracked mobile platforms. This "plug-and-play" design reduces setup time significantly compared to traditional integration methods.
Industry Insight
- Vertical Specialization is Key to Embodied AI Success: General-purpose humanoid robots may face longer adoption curves in heavy industry; instead, specialized agents solving specific, high-pain-point tasks (like welding in non-standard environments) offer faster ROI and clearer market entry strategies.
- Shift from CapEx to OpEx Models: The exploration of "robot labor" services (paying per meter welded) suggests a future where manufacturers prefer operational expenditure models over capital investment, lowering the barrier to entry for adopting advanced robotics in SMEs.
- Supply Chain Autonomy as a Competitive Advantage: Companies that control both the AI algorithms and the hardware production lines (as 3Srobotics does) can iterate faster and ensure quality consistency, distinguishing themselves from pure software solution providers who rely on third-party hardware integrations.
Disclaimer: The above content is generated by AI and is for reference only.