Robots can see with cameras and hear with microphones, but most of them cannot feel. That single missing sense is why a robot will crush an egg or drop a glass while a toddler handles both without thinking. A new result aims at the reason touch has stayed rare on robots, and the reason is not science but cost: a team has built a flexible fibre that senses both stretch and pressure and is cheap and simple to manufacture.
The work was posted this week as a preprint, meaning a study shared publicly before other scientists have formally reviewed it, so it should be read as a promising early result rather than a settled one.
Here is the problem it addresses. Giving a robot a sense of touch means covering it in sensors, the way your skin is covered in nerve endings. The technology to do this exists, but it has been expensive, delicate, and fiddly to make, which is why touch sensing on robots is usually limited to a fingertip or two rather than spread across a whole body. A robot that can only feel with its fingertips is like a person whose body is numb everywhere except the tips of two fingers.
The new fibre attacks the cost and manufacturing side of that bottleneck. It is thin and bendable, so it can be woven into a covering or wrapped around a robot’s limbs, and it reports two different things at once: how much it is being stretched, and where and how hard it is being touched. Stretch sensing tells a soft robot what shape it is currently in. Touch sensing tells it when, and how firmly, it has made contact with the world. Crucially, the design prioritises being inexpensive and easy to produce, which is the property that decides whether a sensor stays in the lab or ends up on real machines.
What does this change? If touch sensing becomes cheap enough to cover an entire robot rather than dot a fingertip, the kinds of jobs robots can attempt widen considerably. Whole-body touch is what lets a machine handle fragile or oddly shaped objects, work safely shoulder to shoulder with people, and notice immediately when it bumps something it should not. The same idea reaches beyond robots, into prosthetic limbs that could restore a sense of touch and into wearable garments that read body movement. In every case the barrier has been the same: not whether it can be done, but whether it can be done affordably and at scale. Lowering that cost is the quiet unlock.
This also fits a larger shift in robotics away from rigid metal arms toward soft robots, machines built from flexible materials that bend and squeeze rather than clank. Soft robots need soft senses, and a stretchy fibre that feels is a far better match for a stretchy body than a hard electronic chip.
The limits are worth stating plainly. This is a single preprint, not yet peer reviewed, and a sensor that performs well in a lab still has to prove it can survive thousands of bends, washes, and knocks in the real world, integrate cleanly with a robot’s electronics, and actually be manufactured cheaply at volume rather than just in principle. Those are the questions that decide whether this becomes a product.
If the cost claims hold up, though, the significance is real. The thing standing between today’s clumsy robots and machines that handle the world gently has often been the price of touch. This is an attempt to bring that price down.
Sources
Ray Jackson holds a BSc in Electrical Engineering from the University of Manitoba and a PhD in Physics from Carleton University. His reporting interests include Current and Future Technologies, Engineering and Artificial Intelligence.