汽车座椅行业的同仁们对祥鑫科技应该不陌生。作为一家以精密冲压模具和金属结构件为主业的上市公司，祥鑫科技在汽车座椅骨架领域扮演着Tier 2供应商的角色，通过间接方式向国内等主流车企供货，并确认为江淮“尊界”系列供应座椅骨架产品。近年来，祥鑫科技将人形机器人定位为“未来发展战略的核心方向之一”。
Colleagues in the automotive seat industry should be familiar with Xiangxin Technology. As a listed company with precision stamping dies and metal structural parts as its main business, Xiangxin Technology plays a Tier 2 supplier role in the automotive seat frame field, supplying domestic mainstream automakers indirectly, and has confirmed supplying seat frame products for JAC's "Zunjie" series. In recent years, Xiangxin Technology has positioned humanoid robots as "one of the core directions of future development strategy."
2026年10月12日，祥鑫科技参股的合肥第四定律机器人有限公司宣布，其自主研发的具身智能模型Axiom-0在具身智能评测基准RoboCasa365 Benchmark中取得公开排行榜第一。这家公司成立于2026年6月，成立仅三个多月。8月，祥鑫科技即对其进行千万级种子轮投资。Axiom-0模型的定位是面向复杂机器人任务场景，通过创新机制提升机器人在多任务环境中的泛化能力。双方将围绕汽车制造、智能装配等真实工业场景开展合作探索，推动具身智能机器人从仿真环境走向工业生产现场。对于汽车座椅行业而言，这意味着具身智能技术有望在座椅装配线等场景中率先落地。
On October 12, 2026, Hefei Fourth Law Robotics Co., Ltd., in which Xiangxin Technology holds a stake, announced that its self-developed embodied intelligence model Axiom-0 ranked first on the public leaderboard of the embodied intelligence benchmark RoboCasa365 Benchmark. This company was founded in June 2026, only over three months old. In August, Xiangxin Technology made a ten-million-yuan seed round investment in it. The Axiom-0 model is positioned for complex robot task scenarios, improving robots' generalization capability in multi-task environments through innovative mechanisms. The two parties will cooperate on real industrial scenarios including automotive manufacturing and intelligent assembly, advancing embodied intelligence robots from simulation environments to industrial production sites. For the automotive seat industry, this means embodied intelligence technology is expected to land first in scenarios like seat assembly lines.
灵巧手是祥鑫科技在机器人零部件领域的拳头产品。经过近两年研发，公司已完成两代灵巧手产品的迭代开发。根据公司披露的信息，第二代灵巧手的技术参数如下。轻量化设计方面，整手重量仅725g，抓取力度达15N。感知能力方面，全掌布置16个触觉传感器模块，510个掌心阵列点可实时感知0.1至15N的压力变化和物体表面软硬程度。精度指标方面，指尖重复定位精度±0.09mm，控制频率不低于500Hz，通讯速率达1KHz。耐用性方面，通过260万次以上耐用性测试。环境适应性方面，在-20℃至60℃宽温范围内稳定工作，可应对焊接车间高温和冷链仓储低温等极端工况。
The dexterous hand is Xiangxin Technology's flagship product in robot components. After nearly two years of R&D, the company has completed iterative development of two generations of dexterous hand products. According to company disclosures, the second-generation dexterous hand's technical parameters are as follows. On lightweight design, the entire hand weighs only 725g, with grip force reaching 15N. On perception, 16 tactile sensor modules are arranged across the palm, with 510 palm array points capable of real-time sensing of 0.1–15N pressure changes and object surface hardness. On precision metrics, fingertip repeat positioning accuracy is ±0.09mm, control frequency not less than 500Hz, communication rate reaching 1KHz. On durability, it passes over 2.6 million durability tests. On environmental adaptability, it operates stably across -20°C to 60°C, handling extreme conditions like welding workshop high temperatures and cold chain warehouse low temperatures.
产业化方面，公司已完成灵巧手专用产线的全面建设，具备年产3万只的生产能力，目前已与国内外机器人头部企业达成深度合作，产品进入样品送样和小批量交付阶段。祥鑫科技这条路径，和前面讨论的巨蟹智能有相似之处。两家都是从汽车产业链切入机器人零部件。巨蟹智能从汽车电机的供应链经验出发，做机器人关节模组，祥鑫科技从汽车座椅骨架的精密冲压和结构件经验出发，做灵巧手。汽车行业对零部件的精度、可靠性、成本控制有严苛要求，这套能力迁移到机器人零部件上，构成了一种先发优势。汽车Tier 2供应商的毛利率通常不高，但规模化和成本控制能力强。祥鑫科技做灵巧手，能不能把这种能力复用到机器人赛道，是它接下来要回答的问题。
On industrialization, the company has completed comprehensive construction of a dedicated dexterous hand production line, with annual capacity of 30,000 units. It has reached deep cooperation with leading domestic and international robotics companies, with products entering sample delivery and small-batch delivery stages. Xiangxin Technology's path resembles Juxie Intelligent discussed earlier. Both enter robot components from the automotive industry chain. Juxie Intelligent starts from automotive motor supply chain experience to make robot joint modules; Xiangxin Technology starts from automotive seat frame precision stamping and structural parts experience to make dexterous hands. The automotive industry has stringent requirements for component precision, reliability, and cost control. Migrating this capability to robot components constitutes a first-mover advantage. Automotive Tier 2 suppliers typically have low gross margins but strong scaling and cost control capabilities. Whether Xiangxin Technology can reuse this capability in the robot track when making dexterous hands is the question it must answer next.
