Next-Gen Robotic Skin Brings Touch Closer to Human Sensation
Scientists from the University of Cambridge and University College London (UCL) have developed a low-cost, highly sensitive, and flexible robotic skin that could revolutionize how machines interact with their environments. Designed to wrap around robotic hands like a glove, this new skin allows robots to sense pressure, temperature, and even damage — mimicking the functionality of human skin more closely than ever before.
Unlike traditional robotic touch systems that rely on a collection of separate sensors embedded into soft materials, this breakthrough material acts as a single unified sensor. The entire surface of the electronic skin responds to different stimuli, much like human skin, enabling robots to detect and interpret multiple forms of contact using just one type of sensor. This approach is called multi-modal sensing.
Crafted from a gelatin-based hydrogel that is soft, stretchy, and electrically conductive, the skin can be melted down, reshaped, and re-cast into complex forms like a human hand. Researchers placed just 32 electrodes at the wrist, but thanks to the material’s 860,000+ internal pathways, the system could capture over 1.7 million individual signals, offering incredibly detailed touch mapping.
From Scalpel Cuts to Heat Guns: Teaching Robots to Feel
To train the skin to recognize different types of touch, the team used a range of physical tests and machine learning algorithms. They blasted it with heat, poked and pressed it with fingers and robotic arms, and even sliced it with a scalpel — all while collecting massive datasets of sensor responses.
This data was then used to train AI models capable of interpreting what kind of touch was being experienced — from a gentle tap to severe damage. Over time, the robotic skin could not only detect where and how it was touched, but understand the nature of the interaction.
“Instead of needing one sensor for heat and another for pressure, we’re using one material that responds to everything, and letting the AI figure out what it means,” explained lead author Dr. David Hardman, a postdoctoral researcher in Professor Fumiya Iida’s lab at Cambridge. Co-author Dr. Thomas George Thuruthel from UCL added that this method is cheaper, easier to fabricate, and more robust than traditional sensor arrays.
While this artificial skin doesn’t yet match the sensitivity of human skin, researchers believe it surpasses any current alternatives, with promising applications in humanoid robotics, prosthetics, automotive systems, and even disaster relief robots. The team now plans to improve durability and conduct real-world testing in robotic applications.
The project was supported by the Samsung Global Research Outreach Program, the Royal Society, and the Engineering and Physical Sciences Research Council (EPSRC) under UK Research and Innovation (UKRI).
