The material's properties remain stable across repeated use
For years, the promise of soft, flexible sensors has been shadowed by a quiet flaw: press them once, and they answer differently than they did before. Researchers at the National University of Singapore have now confronted this fundamental unreliability, engineering a material called TRACE that holds its accuracy across repeated use by drawing on the physics of how surfaces meet. Published in the Proceedings of the National Academy of Sciences in late 2020, their work does not merely improve a device — it removes a barrier that has kept soft electronics from earning trust in medicine, robotics, and the care of the human body.
- Every soft sensor ever built has carried a hidden flaw: the material shifts between presses, making each reading slightly less trustworthy than the last — a problem serious enough to render wearable health data unreliable.
- NUS researchers cracked metal films into precise ring-shaped patterns and fused them to a flexible polymer, creating a composite that resists the drift other soft materials cannot escape.
- In repeated stress tests, TRACE delivered five times greater consistency than conventional soft sensor materials, a margin large enough to move the technology from laboratory curiosity to practical tool.
- The implications ripple outward — toward wristbands that accurately track arterial blood flow, robotic hands that can genuinely feel texture, and prosthetic limbs that respond to the world with something closer to natural sensation.
- The team's horizon is a skin-adhering smart patch that pairs TRACE sensors with machine learning to read cardiovascular patterns continuously — a future that now has a credible foundation beneath it.
Soft materials move with the body in ways rigid electronics never can, making them the natural choice for wearables and robotic skin. But flexibility has always carried a hidden cost: press a soft sensor repeatedly, and the material itself shifts, returning slightly different readings each time. Scientists call this drift hysteresis, and it has quietly blocked soft sensors from earning a place in real health monitoring or reliable robotics.
Assistant Professor Benjamin Tee and his team at the National University of Singapore decided to attack the problem at its physical root, drawing on contact mechanics — the science of how surfaces press against one another. What they produced is TRACE, or Tactile Resistive Annularly Cracked E-Skin, built by fracturing metal thin films into ring-shaped patterns and bonding them to a flexible polymer substrate. When tested under repeated mechanical stress, TRACE maintained consistent readings at five times the reliability of conventional soft sensor materials — a result significant enough to appear in the Proceedings of the National Academy of Sciences in September 2020.
The stakes are practical and immediate. A health wearable that gives a different reading each time it is worn is not a medical instrument — it is noise. A robotic hand that cannot feel texture consistently cannot learn to handle objects safely. TRACE addresses both failures by keeping its material properties stable across use, turning soft sensing from an unreliable novelty into something that can be trusted.
The team envisions three converging futures for the technology: wearables that measure arterial blood flow near the skin's surface for cardiovascular monitoring, robotic systems with genuine tactile intelligence, and prosthetic limbs with skin interfaces responsive enough to feel natural. Further ahead, they are working toward a miniature smart patch that combines TRACE sensors with machine learning to detect cardiovascular patterns from pulse velocity data — reading health continuously from the surface of the skin. The materials problem has been solved; the door it opens is only beginning to be walked through.
Soft materials bend and flex in ways rigid electronics cannot, which makes them ideal for wearables that need to move with the body and robots that need to sense their environment. But flexibility comes with a cost: these materials are unreliable. Every time you press on them, they respond slightly differently. The material itself changes shape between one touch and the next, introducing error into the readings. Scientists call this drift hysteresis, and it has been the stubborn problem holding back soft sensors from real-world use in health monitoring and robotics.
A team at the National University of Singapore, led by Assistant Professor Benjamin Tee, set out to solve this. Their approach was rooted in contact mechanics—the physics of how surfaces press against each other. What they built is a sensor material that maintains its accuracy across repeated compressions, something conventional soft electronics have struggled to do. In September 2020, they published their results in the Proceedings of the National Academy of Sciences.
The innovation is called TRACE, short for Tactile Resistive Annularly Cracked E-Skin. The team created it by cracking metal thin films into ring-shaped patterns and bonding them to a flexible polymer called polydimethylsiloxane, or PDMS. They then integrated this composite with electrodes and substrates to build a pressure-sensing device. When they tested it repeatedly under mechanical stress, the results were striking: TRACE performed five times better than conventional soft sensor materials at maintaining consistent readings.
What makes this matter is accuracy. A wearable health device that gives you a different reading every time you put it on is not useful. A robotic hand that cannot reliably feel the texture of what it is touching cannot learn to manipulate objects safely. The hysteresis problem meant that soft sensors could detect pressure, but the data they returned was noisy and unreliable. TRACE changes that equation. The material's properties remain stable across repeated use, which means the sensor can deliver trustworthy measurements every time.
Tee and his team see multiple paths forward. One is wearable health monitoring—specifically, measuring blood flow in arteries near the skin's surface to track cardiovascular health. Another is robotics, where reliable tactile sensing could help machines perceive surface texture and respond intelligently to what they touch. A third application is prosthetics, where a dependable skin interface could allow artificial limbs to respond more naturally to the world around them.
The team's longer-term vision is more ambitious. They want to develop a tiny smart patch that adheres to human skin and uses TRACE sensors combined with machine learning to predict cardiovascular health. The sensor would capture pulse velocity data with high fidelity, and algorithms would learn to recognize patterns that signal health or disease. That is still ahead, but the foundation is now in place. What began as a materials science problem—how to make soft sensors reliable—has opened a door to wearable devices that could fundamentally change how we monitor our own bodies.
Citas Notables
Our long-term goal is to predict cardiovascular health in the form of a tiny smart patch that is placed on human skin.— Assistant Professor Benjamin Tee, NUS
La Conversación del Hearth Otra perspectiva de la historia
Why does flexibility make these sensors less reliable? It seems like it should be the opposite.
When a soft material gets pressed, it deforms. But when you release the pressure, it doesn't always return to exactly the same shape. The material's internal structure shifts slightly. So the next time you press it, the electrical properties are a little different. That variation—that drift—is hysteresis. It's the material's memory of being touched.
And this TRACE material solves that by doing what, exactly?
By cracking the metal film into rings. Those cracks create a specific pattern that distributes stress more evenly across the material. The geometry itself becomes part of the solution. It's not just about the materials; it's about how they're arranged.
Five times better is a big jump. What does that actually mean in practice?
It means if you measure your pulse ten times in a row with a conventional soft sensor, you might get ten slightly different numbers. With TRACE, those numbers stay consistent. For health monitoring, that consistency is everything. You need to trust what the device is telling you.
The smart patch idea—how far away is that?
The sensor itself is ready. What comes next is integrating it with AI and making sure the patch can stay on skin safely for extended periods. They're working on conformability now, making sure it can bend and stretch with the body without losing its sensing ability. That's the next hurdle.
And prosthetics—why does this matter there?
A prosthetic limb needs to feel what it's touching so it can respond appropriately. If the sensors are unreliable, the limb can't learn. With TRACE, an artificial hand could actually develop a sense of texture, pressure, and grip strength. It becomes more than a mechanical replacement; it becomes something that can adapt.