Robotiq has launched its TSF-85 tactile sensor fingertips for the 2F-85 Adaptive Gripper, giving Physical AI systems the sense of touch needed for real-world manipulation.
With integrated tactile sensing, Robotiq enables robots to not only see the world, but to feel, understand, and reliably interact with it at scale. The Tactile Sensor Fingertips for the 2F-85 gripper are available now.
Robots cannot learn the physical world through vision alone. Real-world manipulation requires contact awareness, force control, and feedback from interaction.
With integrated tactile sensing, Robotiq expands its proven manipulation system beyond vision alone, enabling robots to feel, understand, and reliably interact with physical objects at scale. The solution is designed for immediate use in AI training labs, humanoid robotics, and industrial physical AI applications.
Vincent Duchaine, CTO artificial intelligence at Robotiq, says: “Physical AI demands more than clever algorithms – it demands reliable interaction with the real world.
“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.”
Adaptive gripping meets touch-enabled dexterity
Robotiq’s 2F-85 Adaptive Gripper is built on a patented mechanical design that goes beyond a traditional parallel gripper.
While parallel grippers rely on precise positioning and rigid alignment, the 2F gripper provides inherent adaptability by offering both pinch and encompassing grips, with stroke lengths of 85 mm and 140 mm.
This allows the gripper to conform to object shape, reduce grasp planning complexity, and minimize reliance on perfect vision, making it well suited for general-purpose physical AI systems designed to handle a wide variety of objects.
The new tactile fingertips add a powerful sensing layer:
- A 4×7 static taxel grid to monitor force distribution
- Micro-slip detection at 1000 Hz for stable, precise manipulation
- An integrated IMU for proprioceptive sensing and contact awareness
Together, these capabilities allow robots to understand contact geometry, detect incipient slip, and improve generalization across diverse objects, capabilities that are critical for physical AI datasets.
Built for fleets, not just demos
Unlike fragile, custom-built tactile hands that often struggle with durability and operational limits, Robotiq’s solutions are designed for scalable, long-term deployment.
Thousands of Robotiq grippers are already operating globally in demanding industrial and research environments, delivering high uptime, consistent performance, and low total cost of ownership.
The tactile-enabled 2F grippers integrate easily into existing systems using native RS-485 communication and a USB conversion board, enabling flexible deployment across robot brands and research platforms.
The tactile fingertips are designed to preserve the gripper’s encompassing and pinch grip mechanics, with minimal impact on stroke and reach, and feature robust cabling built for real-world operation.
With a lower bill of materials and replacement cost than anthropomorphic or DIY hands, Robotiq provides a practical path from lab prototype to large-scale robot fleets.
Physical AI-ready, from training to deployment
Robotiq supports modern AI workflows by delivering reliable, feature-dense hardware designed to work consistently from proof of concept through real-world deployment.
By providing stable sensing and repeatable interactions across environments, Robotiq offers a reliable base for manipulation for reinforcement learning (RL), vision-language-action (VLA) models, and imitation learning.
The TSF-85 tactile sensor fingertips build on years of Robotiq research and field experience. Robotiq shares best practices for tactile data handling, including guidance on bias management, normalization, and outlier detection, to help teams generate consistent, high-quality data for model training.
By standardizing hardware and tactile data across robot fleets and research campuses, Robotiq reduces integration friction and accelerates experimentation. This reduced integration helps teams move faster from lab validation to production-scale physical AI systems.
Proven by the world’s leading AI and robotics teams
The TSF-85 joins Robotiq’s portfolio of physical AI-ready manipulation components, building on a foundation already proven at scale.
With more than 23,000 grippers deployed worldwide, Robotiq’s technology is trusted by leading organizations and has demonstrated strong performance in AI training labs and humanoid robotics pilots.
Aleksei Filippov, head of business development at Yango Tech Robotics, says: “To build physical AI that truly works, you need hardware that can sense, respond, and learn from every interaction.
“That’s why we chose Robotiq. With Robotiq precision force control and reliable feedback, we capture rich sensory data from every grasp.”
Compared to do-it-yourself (DIY) tactile hands that take months to develop and maintain, Robotiq delivers a ready-today, field-proven solution. And compared to anthropomorphic hands that are expensive and complex, Robotiq achieves the majority of real-world manipulation tasks at a fraction of the cost.
Enabling the future of physical AI
By giving robots a reliable sense of touch that combines with adaptive gripping and industrial-grade reliability, Robotiq enables the next generation of physical AI systems to learn faster, operate more robustly, and scale beyond the lab.
Backed by a large network of local partners, Robotiq ensures that customers have fast support for setup and ongoing service. This support helps teams get back up and running quickly if challenges arise.
Main image: The tactile fingertips allows robots to understand contact geometry, detect incipient slip, and improve generalization across diverse objects, capabilities that are critical for physical AI datasets
