CATL Backs Humanoid Tactile Sensor Firm Link-touch

In a significant move underscoring the growing importance of sensory feedback in robotics, CATL Capital, the investment arm of global battery giant CATL, has led a substantial funding round for Link-touch (also known as Lan Dian Chuzhi). The Chinese startup, a leader in six-axis force-torque sensors for humanoid robots, secured over 100 million yuan (approximately $13.8 million USD) in an extended Series C financing. This marks the third major funding injection for Link-touch within the past year, highlighting rapid investor confidence in its technology.

The strategic investment, which also saw participation from prominent players like machine vision specialist OPT and robot manufacturers AgiBot and Galbot, will fuel Link-touch's development of next-generation force-sensing technology, accelerate production scaling, and support international market expansion. Link-touch's advanced sensors are crucial for enabling robots to perform delicate manipulation tasks, a key bottleneck in the advancement of embodied AI.

Founded in 2019, Link-touch has established a robust technological foundation across four core platforms: elastic structure design, embedded hardware, decoupling algorithms, and six-axis calibration. This allows them to offer a diverse range of sensors suitable for everything from sophisticated humanoid robots to general industrial applications. Their flagship products boast impressive specifications, including sub-0.1% full-scale error and response frequencies exceeding 10kHz, rivaling established international competitors and positioning them as a dominant force in the domestic market.

Market intelligence from MIR (睿工業) indicates Link-touch commanded approximately 62% of China's six-axis force-torque sensor market for humanoids in 2024, with projections reaching 70% in 2025. Further data from GGII (高工机器人产业研究所) shows their market share stood at 72.6% for the first nine months of 2025, solidifying their leading position. The company has already secured supply agreements with major humanoid developers, including Zhiyuan Robot (智元機器人), Xiaomi, Xpeng, Galbot (銀河通用), and UBTECH.

The involvement of CATL Capital is particularly noteworthy. CATL has been actively investing in the embodied AI sector, extending its reach beyond finished robot products to critical component suppliers like Link-touch. Beyond capital, CATL plans to support Link-touch through its extensive customer network, manufacturing expertise, and potential integration into its own production facilities. The deepened relationships with AgiBot and Galbot, transitioning from customers to shareholders, signal a commitment to collaborative development in areas like precision assembly and force control for grasping.

Link-touch CEO Liu Wuyue emphasizes the critical role of force control for precise robotic movements. While machine vision excels at environmental perception, fine motor skills rely heavily on tactile feedback. The complex, non-linear nature of force data, influenced by object properties and robot arm dynamics, makes it difficult to simulate and necessitates real-world data acquisition. Consequently, force sensors are vital for gathering the training data essential for advancing AI models in robotics.

The global humanoid robot market is experiencing explosive growth, with IDC estimating shipments to reach approximately 18,000 units in 2025, a more than 500% increase year-over-year. Morgan Stanley has significantly revised its forecasts for China's humanoid market, projecting sales between 14,000 and 28,000 units in 2026 and a staggering 262,000 units by 2030. In this dynamic environment, Link-touch has achieved triple-digit revenue growth for three consecutive years and reached profitability. The company anticipates shipping around 400,000 joint force sensors and 30,000-40,000 six-axis force-torque sensors in 2026. Their new 10,000+ square meter production facility in Guangdong is nearing full automation, with plans to scale annual capacity to 1 million joint sensors and 200,000 six-axis sensors, reducing lead times to just two weeks.