Researchers at Queen Mary University of London have developed a soft material that changes color when pressed, allowing robots to read touch directly from the light reflected off its surface. The material, developed by postdoctoral researcher Giacomo Sasso, addresses a long-standing tradeoff in robotic tactile sensing between detail and speed. Instead of reconstructing pressure through software analysis, the material displays force as a color pattern that can be captured with a standard low-cost USB camera and interpreted without additional computation.
The color response is generated by the material’s microscopic structure rather than any embedded dye, drawing on structural color effects that have been studied in mechanochromic materials for years. Layered patterns thinner than a hair reflect specific wavelengths of light, and when pressure compresses those layers together, the color at that point shifts. Because the color already encodes the applied force, the pressure map updates as fast as the camera can film it. In tests, the system resolved details including the fine relief of a stamped coin and the ridge pattern of a human fingerprint, a level of detail earlier tactile sensors could not achieve.
The approach differs from established vision-based tactile sensing, in which a camera films a soft gel from behind and software works backward from the deformation to reconstruct the applied force. That reconstruction step introduces latency and typically forces a tradeoff with resolution. Another recent color-based system relies on a trained neural network to interpret its color patterns. Sasso described the Queen Mary approach as moving toward embedded intelligence, with sensing built into the material itself rather than added through downstream software. A single human hand contains more than 10,000 tactile receptors, and traditional attempts to match that density have required dense sensor arrays and extensive wiring.
Practical applications remain at the prototype stage, but the researchers point to precision manufacturing, prosthetics, and surgical robotics as early targets. In assembly of small or fragile components, the material could detect grip pressure changes before damage occurs. In surgical settings, it could translate stiffness differences between healthy tissue and tumors into visible color changes readable by tools or robotic systems. Sasso noted that significant engineering work remains before commercial or clinical deployment. Still, the elimination of a reconstruction step from vision-based touch sensing, combined with low-cost hardware requirements, points to a viable path for higher-resolution tactile feedback in humanoid, surgical, and industrial robotic platforms.
