Additionally learn NVIDIA’s COMPUTEX protection, the place Robotiq seems alongside the most recent Isaac GR00T updates.
Robotiq has launched the digital twin of its TSF-85 tactile sensor in NVIDIA Isaac Sim, the primary industrial-grade tactile sensor digital twin transport on a business collaborative gripper. Tactile sensing guarantees to speed up robotics, however its adoption has been restricted by the shortage of information from industry-ready {hardware} and correct simulations. Mannequin builders can now prepare contact-rich manipulation insurance policies in simulation, then run them on the identical bodily sensor designed to function reliably on the manufacturing facility ground.
Most tactile sensors depend on a deformable contact interface. The very property that offers them sensitivity can be what makes them troublesome to simulate. Deformable physique simulation is technically demanding, and that’s one cause correct tactile digital twins have lagged behind within the Bodily AI stack. The TSF-85 digital twin is constructed round that constraint. It generates artificial tactile maps by a customized Isaac Sim UI panel, visualizes them in actual time, runs information technology on the simulation refresh price, and exports to HDF5 for downstream coaching pipelines.
The TSF-85 digital twin was developed by the CoRo Lab (Laboratoire de commande et de robotique) at École de technologie supérieure (ÉTS) in Montréal, a long-time analysis companion of Robotiq — a collaboration led by Affiliate Professor Jean-Philippe Roberge and doctoral researcher Berith Atemoztli De la Cruz Sánchez.
The simulation technique behind the digital twin is documented in two peer-reviewed publications cited within the GitHub repo. The primary, revealed in Frontiers in Robotics and AI (2025), attracts on a dataset of 53,400 real-world tactile maps to coach, validate, and check every simulation pipeline — reaching as much as 97% Structural Similarity Index Measure (SSIM) for the hyperelastic mannequin and 90% SSIM for the elastic mannequin on 12 unseen objects. A companion paper at ICCRT 2025 releases an open dataset of 46,200 actual and artificial tactile samples, together with information collected utilizing a 2F-85 Robotiq gripper and artificial samples generated in NVIDIA Isaac Lab.
Simulation accuracy issues, but it surely solely helps if the true sensor stays steady over time. The TSF-85 has been examined by 2.3 million cycles at most gripper drive, with no important variation within the output of the tactile indicators. This implies fashions skilled on its tactile information can proceed to depend on constant indicators for edges, shapes, textures, and geometry even after demanding real-world use.
Robotiq has been the go-to elements supplier for each tutorial analysis labs and industrial manufacturing for greater than a decade. That twin footprint is strictly the bridge Bodily AI requires: research-grade flexibility and industrial-grade reliability in the identical {hardware} platform.
The digital twin helps NVIDIA Isaac Sim 5.1 and is obtainable now on GitHub: https://github.com/Lab-CORO/TSF-85. Be taught extra about Robotiq’s bodily AI stack at https://robotiq.com/physical-ai
