Robotiq 2FC Gripper Adds Embedded AI to 30,000-Unit Installed Base of Industrial Grippers
Tactile sensing plus local AI inference give robots real-time grip awareness without the cloud
Robotiq launched the 2FC physical AI gripper on October 22, 2026, opening pre-orders for a device that combines tactile fingertip sensors, a wrist-mounted inference computer, and an embedded software platform into a single production-rated end-effector. Designed for collaborative and humanoid robots, the 2FC is the company's first product to run AI grasp skills directly at the gripper—without cloud connectivity, without an external compute box.
The launch lands into a physical AI market building urgency for years. Robots have long excelled at repeatable pick-and-place operations in controlled environments. What they lack is the ability to adapt a grasp in real time based on what is happening at the point of contact—the slipping of a wet part, the give of a soft bag, the shift of an oddly balanced object. Without that feedback loop, automation struggles against the irregular contact conditions that define industrial assembly, food handling, and logistics at scale.
Six Sensor Modalities at Millisecond Resolution
The 2FC captures six synchronized data streams at millisecond timing: pressure, vibration, proprioception, force, position, and current. The pressure data comes from 28 taxels—individual touch-sensitive elements—embedded in the fingertips, each sampled at 1,000 Hz. The vibration channel enables one of the 2FC's most practically valuable capabilities: incipient slip detection, which identifies micro-vibration signatures that precede a slip event and corrects grip force before an object is dropped.
Embedded Inference at the Wrist, Not in the Cloud
All inference runs locally on the AI Hub, a compute module mounted at the wrist. Rather than streaming raw sensor data upstream for external processing, the 2FC emits labeled events—structured outputs from five embedded grasp skills: grasp quality assessment, incipient slip detection, object pose estimation, object centering, and advanced grip check. The distinction matters in factory deployments where cloud round-trip latency can mean the difference between a caught slip and a dropped part.
Contact Core Turns Sensors Into a Software Platform
The software layer, Contact Core, transforms the 2FC from a sensor-equipped gripper into an open platform. Released as open source ahead of the hardware launch, Contact Core handles synchronized multi-modal data streams, provides simulation assets for NVIDIA Isaac Sim to support sim-to-real training, and ships with ROS 2 packages and a C++ SDK. The Isaac Sim integration allows developers to generate synthetic training data in simulation and transfer resulting models to the physical gripper—a pipeline that matters as companies invest in physically realistic manipulation policies.
A 12-Year Path From Research Bench to Production Floor
The journey from sensing hardware to embedded AI was incremental. The TSF-85 add-on tactile fingertips launched in January 2026 provided sensor hardware for existing 2F-85 grippers but required external compute to process the data. Contact Core followed as open-source software in September 2026. The 2FC integrates all three layers—sensors, on-wrist compute, and software—into one production package rated for two million cycles.
"We've spent decades teaching robots to see the world better," said Samuel Bouchard, CEO of Robotiq, who co-founded the company in Quebec City in 2008 after building robotics expertise at the Laval University Robotics Lab. "Now we're starting to understand how much intelligence comes from what happens when they actually touch it. When we built our first tactile sensors 12 years ago, the problem wasn't sensing. The problem was what to do with all that information. This time, the intelligence needed to make use of touch data is finally catching up, and with the 2FC we are embedding that tactile intelligence directly into our products."
Existing Gripper Users Get an Upgrade Path
Existing 2F-85 and 2F-140 gripper users can upgrade their installations with tactile sensor fingertips and an AI Hub module rather than replacing the full gripper, preserving the installed base of more than 30,000 deployed units.
Where the 2FC Sits in the Tactile Sensing Landscape
Standalone tactile sensors have existed for years—XELA Robotics, Contactile's PapillArray, and vision-based approaches from GelSight and FingerVision each address parts of the problem. What has been largely absent is a production-rated system that integrates sensors, local compute, and actionable AI skills into a single end-effector with commercial lifecycle guarantees. The 2FC is positioned to fill that gap on a gripper platform already deployed across automotive, electronics, and food processing applications.
Pre-Orders Open, Delivery in Q1 2027
The 2FC ships in two variants—2FC-85 and 2FC-140—covering different stroke ranges and part sizes. Pre-orders are open as of today, with initial deliveries planned for Q1 2027. Robotiq is hosting a launch webinar on October 28, 2026, at 11:00 AM ET.
The more consequential near-term test is whether Contact Core's open-source model attracts third-party grasp skill development—the same dynamic that made ROS a standard for robot middleware. The humanoid robot sector, where physical manipulation remains one of the hardest unsolved problems, is paying close attention: Robotiq has explicitly flagged humanoid integration in its roadmap, and with the physical AI market growing at over 47 percent annually, commercial pressure to close the touch gap is substantial.