Tactile Sensing Sector Booms as Chinese Sensor Firms Accelerate Commercialization

As humanoid robots move from exhibition booths to factory floors, tactile sensing—a technology once overlooked—is becoming a focal point of industrial competition.

Since the start of 2026, total funding in the tactile sensing sector has surpassed RMB 7 billion, while the penetration rate of tactile sensors in commercial humanoid dexterous hands has climbed from 20% at the beginning of the year to more than 60%.

That growth rate is rare among robot component categories, underscoring how quickly demand for tactile perception is rising as robots enter real work environments.

The surge is not merely a capital-market phenomenon. It reflects a broader shift in robotics: after years of progress in vision-language-action models, the industry is confronting tasks where sight alone cannot ensure reliable manipulation.

Precision assembly and flexible grasping require force control and slip detection, capabilities that tactile sensing provides. For robots to move beyond demonstrations and into production lines, touch is becoming as important as vision.

The momentum suggests tactile sensing is moving from an optional add-on to a core component, especially as humanoid robot commercialization accelerates. Factories need robots that can handle delicate parts without damaging them, and tactile feedback is central to that reliability.

Qianjue Robotics recently announced a funding round in the hundreds of millions of RMB, led by strategic investors from the embodied intelligence industry and Jide Electric.

Founded in May 2024, the company was started by Ma Daolin, an associate professor at Shanghai Jiao Tong University who received the only Best Paper Award at ICRA 2021.

Qianjue has built its technology system around multimodal tactile sensors, data collection systems, and a VTLA embodied model. It says it now serves more than 300 leading industry clients. The company’s technology stack aims to connect sensing hardware with embodied intelligence, so that tactile data can directly inform robot control.

Behind the rising demand for tactile perception is a substantive change in robot applications. Vision-language-action models have given robots environmental understanding and task planning, but in precision assembly and flexible grasping, vision alone cannot manage force control or slip detection.

Ma Daolin points out that repetitive, low-precision tasks do not require tactile support, but all high-precision, flexible industrial operations make tactile sensing a necessity. That distinction is shaping where tactile sensors are deployed first.

The evolution of technical routes supports this judgment. Qianjue’s VTLA model incorporates tactile modality into the traditional vision-language-action architecture, enabling the robot to perceive deformation, texture, and force distribution in real time during contact.

Its whole-hand tactile solution extends sensing from fingertips to palm and finger pads. Fingertip sensor marker density reaches 125 points per square centimeter, with resultant force accuracy of 0.03 N. These specifications are aimed at tasks that demand delicate force regulation. The funding follows a wave of investor interest in embodied AI components, where tactile sensing is seen as both a bottleneck and an opportunity. Qianjue’s clients span industrial and service robotics, according to the company, indicating demand beyond a single application niche.

Building data infrastructure is equally critical. Qianjue has introduced a wearable tactile data collection device that synchronously captures vision, touch, pose, and IMU data during natural human-robot interaction.

The device balances cost, efficiency, and quality in data collection. This scalable data-gathering capability provides a foundation for training tactile large models. By aligning multisensory data streams, such systems aim to reduce the cost of collecting high-quality tactile datasets at scale.

From an industry perspective, the tactile sensor sector is still in a stage of diverging technical routes. MEMS, flexible electronics, and vision-based tactile sensing each have different strengths, and the market has not yet produced a winner-take-all outcome.

What is certain is that as hardware costs fall and data accumulation accelerates, tactile perception is shifting from an optional robot configuration to an essential capability for entering real work scenarios. That transition will likely determine which sensor makers can move from prototypes to mass deployment in the coming years. For now, competition is open, but the companies that solve data and cost challenges may set the standards for tactile-enabled robots.