A team of engineers has developed artificial skin that gives robots a near-human sense of touch. The skin, made from a flexible gel material, can transform a robotic hand's surface into a sensitive, intelligent sensor.
This new material can detect pressure, temperature, pain, and distinguish multiple contacts simultaneously, unlike other artificial skins that combine different sensors. The robotic skin, developed by the University of Cambridge and University College London, was incorporated like a glove, demonstrating its flexibility and ability to mold into complex shapes.
The technology detects and processes various physical signals, enabling robots to interact better with the physical world. "We wanted to develop a solution that could detect multiple types of touch at once, but in a single material," said David Hardman, lead author of the study.
The skin is composed of an elastic and conductive hydrogel that acts as a complete sensor in itself. Instead of requiring multiple separate sensors, this material detects various stimuli—such as pressure, temperature, or physical damage—from any point on its surface. This is possible thanks to an internal network of over 860,000 microscopic pathways through which electrical signals are transmitted.
To read these signals, researchers placed only 32 electrodes at the base of a robotic hand molded with this material, which was enough to collect more than 1.7 million data points. Then, using machine learning techniques, they trained a model capable of accurately identifying the type of contact received: from a soft touch to a cut or a heat source.
This innovative skin could transform multiple industries by giving robots a perception closer to human. In the field of humanoid robotics and prosthetics, it would allow the development of artificial limbs capable of detecting environmental stimuli with great precision, improving interaction with people and objects. In medical or assistance scenarios, this technology would be key to creating devices that respond to touch or external conditions, making tasks that require delicacy or sensitivity more natural.
Its potential applications also extend to sectors such as automotive and disaster rescue. Robots equipped with this skin could identify hot, dangerous, or damaged surfaces in real-time, which is crucial in risk environments where contact with certain materials can compromise safety. Thanks to its low cost, durability, and ability to adapt to complex shapes, the skin could also be used in the manufacture of industrial robots that are more efficient and sensitive to their environment.