Subscribe

Latest Blog Post

How tactile sensing improves model performance

Vision-language-action models are the current state of the art in robotic manipulation. They still cannot pick up a potato chip without crushing it.

That is the result published earlier this year by the team behind the Video Tactile Action Model (VTAM). On a potato chip pick-and-place task — a task that demands high-fidelity force awareness, where vision alone cannot distinguish a crushing grasp from a holding one — VTAM outperformed the π0.5 baseline by 80%. Across the broader contact-rich benchmark suite, VTAM held a 90% average success rate.¹

The chip is an adversarial example, and that is precisely why it is the right test. At the point of grasp, only contact dynamics carry useful signals. Pressure, vibration, and force/torque tell the policy what is happening, correcting the visual estimation errors that vision-only models cannot detect on their own. A camera, however high its resolution, cannot do that work.

Jennifer Kwiatkowski
By Jennifer Kwiatkowski
on May 07, 2026 in Physical AI. 4 min read time
How tactile sensing improves model performance

Vision-language-action models are the current state of the art in robotic manipulation. They still cannot pick up a potato chip...

Jennifer Kwiatkowski
By Jennifer Kwiatkowski
on May 07, 2026
Read more 4 min read time
What is the best palletizing option for your operation?

The best palletizing solution depends on your production volume, budget, available space, and need for flexibility.

You can go...

Marc Giguère
By Marc Giguère
on May 05, 2026
Read more 5 min read time
Vision-only manipulation is hitting a wall

In 2016, I said something that went against where robotics was heading at the time: vision alone doesn’t work for grasping.

Not...

Vincent Duchaine
By Vincent Duchaine
on Apr 30, 2026
Read more 3 min read time
How Medra built the largest autonomous lab in the United States

Medra Lab 001 is the largest autonomous AI-driven laboratory in the United States, operating continuously with robotics, AI,...

Marc Giguère
By Marc Giguère
on Apr 28, 2026
Read more 4 min read time
Why pharmaceutical manufacturers are standardizing robotic palletizing

Pharmaceutical manufacturers are under pressure to increase output, maintain strict compliance, and protect their workforce,...

Linnea Bruce
By Linnea Bruce
on Apr 23, 2026
Read more 4 min read time
Why Physical AI isn't scaling yet, and what's holding it back

Physical AI is advancing quickly.

AI models can now recognize objects, plan actions, and adapt to new tasks. But despite this...

Linnea Bruce
By Linnea Bruce
on Apr 21, 2026
Read more 5 min read time
AI can decide. But can it act? The missing layer in Physical AI

Artificial intelligence has made impressive progress.

Models can classify images, generate text, and even plan complex...

Louis-Alexis Demers
By Louis-Alexis Demers
on Apr 16, 2026
Read more 5 min read time
How TIDI Products increased palletizing productivity by 30% with automation

TIDI Products, a global manufacturer of infection prevention and patient safety products, transformed its end-of-line...

Linnea Bruce
By Linnea Bruce
on Apr 14, 2026
Read more 4 min read time
Scaling Physical AI: Why grippers and sensors matter for real-world robotics

Physical AI is evolving quickly.

From imitation learning to foundation models, robotics teams are making real progress toward...

Marc Giguère
By Marc Giguère
on Apr 09, 2026
Read more 6 min read time

Leave a comment