Robotiq has launched its TSF-85 tactile sensor fingertips for the 2F-85 Adaptive Gripper, adding integrated tactile sensing capabilities to its existing gripping system. The product is intended to support physical artificial intelligence applications that require reliable contact awareness and force feedback for real-world object manipulation.
The tactile sensor fingertips are designed to extend robotic manipulation beyond vision-based systems by providing direct feedback from physical interaction. According to the company, the system enables robots to detect contact conditions, monitor force distribution, and respond to slip during manipulation, supporting more consistent performance across different objects and environments. The tactile fingertips are available for immediate deployment and are aimed at use in AI training environments, humanoid robotics research, and industrial automation.
The TSF-85 fingertips incorporate a static tactile sensing array arranged in a 4 by 7 grid, along with high-frequency micro-slip detection operating at 1,000 hertz. An integrated inertial measurement unit is included to provide additional proprioceptive and contact awareness data. Together, these sensing elements are intended to allow robots to infer contact geometry and adjust grip behavior during handling tasks.
The tactile components are designed to integrate with Robotiq’s 2F-85 Adaptive Gripper, which uses a mechanical architecture that supports both pinch and encompassing grasp modes. The gripper offers stroke lengths of 85 millimeters and 140 millimeters, enabling it to conform to a range of object shapes and reducing dependence on precise visual alignment. The company positions this approach as an alternative to more complex anthropomorphic robotic hands.
“Physical AI demands more than clever algorithms—it demands reliable interaction with the real world,” said Vincent Duchaine, chief technology officer for artificial intelligence at Robotiq. “By combining adaptive gripping with high-frequency tactile sensing, we’re giving robots the sense of touch and control they need to generalize across objects, tasks, and environments without the cost and complexity of anthropomorphic hands.”
Robotiq stated that the tactile-enabled grippers are designed for durability and long-term deployment rather than short-term demonstrations. The system uses native RS-485 communication with an optional USB conversion board, allowing integration across different robot platforms. The tactile fingertips are engineered to maintain the gripper’s existing grasp mechanics while adding sensing capability, and the cabling is designed for industrial operating conditions.
The company said the hardware is intended to support machine learning workflows, including reinforcement learning, vision-language-action models, and imitation learning, by providing consistent and repeatable tactile data across deployments. Robotiq also provides guidance on handling tactile data, including bias management and normalization, to support the generation of training datasets.
Robotiq reported that more than 23,000 of its grippers are currently deployed worldwide across industrial and research settings. The TSF-85 tactile sensor fingertips are positioned as an extension of this installed base, with the goal of enabling scalable physical AI systems that can transition from laboratory research to production environments.
