Can Robots Finally Feel What They're Holding? Two New Sensors Suggest We're Getting Closer
A pair of new research papers tackle one of robotics' oldest unsolved problems: giving robots a decent sense of touch. The approaches couldn't be more different.
By
What would it actually take for a robot to handle a ripe tomato without crushing it?
I keep coming back to this question. It sounds trivial, almost silly. But it's the kind of thing that exposes exactly how far humanoid robots still have to go. Humans do it without thinking. Robots, even very expensive ones, still struggle. The core issue is touch. Specifically, the lack of it.
Two new papers out of arXiv this week are taking very different swings at the problem, and honestly, I find both of them interesting for different reasons.
A Sensor That Knows When It's Slipping
The first comes from a team working on in-hand manipulation. Their paper, published on arXiv, describes a tactile sensor that combines velocity tracking, force and torque measurement, and a full pressure map, all inside a single deformable contact pad. They call it slip-aware, meaning it can detect when an object is starting to slide before it actually falls.
That last part matters more than it might sound. Most manipulation failures don't happen all at once. They happen because a grip is slowly losing purchase and the robot has no idea. By the time the object moves, it's too late. A sensor that catches the early warning signs of slip is a genuinely useful thing.
The team claims this is the first sensor to combine all these sensing modes in a single compliant structure. I initially thought that seemed like a stretch, but reading the paper, the claim appears to hold up, at least within the specific framing they're using. They're also leaning into practical fabrication: standard PCB manufacturing plus rapid prototyping, which keeps costs down and makes it easier to actually build and iterate on.
Related coverage
More in Humanoids
The headlines are celebrating a $2.5B humanoid robotics deal. I'd pump the brakes a little.
Mark Kowalski · 25 Jun · 6 min
Sometimes the sources don't pan out. Here's what happened when I tried to write a humanoids story this week and ended up with Samsung deals instead.
Sarah Williams · 25 Jun · 3 min
Diffusion models are getting good at imagining robot movements, but 'imaginable' and 'physically possible' aren't the same thing. Researchers are starting to close that gap.
Sarah Williams · 25 Jun · 6 min
A batch of fresh robotics research tackles the same underlying problem from different angles: robots that can see but don't really understand where things are.
