According to a study published in Science Robotics, scientists from the University of Cambridge and University College London (UCL) have developed a low-cost, durable, and highly sensitive robotic ‘skin’ that can be attached to robotic hands like a glove, allowing robots to detect information about their surroundings in a manner similar to humans.
Scientists have developed a low-cost, durable, highly sensitive robotic ‘skin’ that can be added to robotic hands like a glove, enabling robots to detect information about their surroundings in a way that’s similar to humans. Image Credit: University of Cambridge
The researchers created the flexible, conductive skin, which is simple to make and can be melted down and reshaped into various complex shapes. The technology detects and interprets various physical inputs, enabling robots to interact with the real environment in more meaningful ways.
Unlike existing robotic touch solutions, which often use sensors placed in limited regions and require multiple sensors to detect different forms of touch, the Cambridge and UCL researchers' electronic skin comprises sensors, bringing it closer to the skin.
Even though the robotic skin is not as sensitive as human skin, it can identify various forms of touch and pressure in a single material, such as a finger tapping, a hot or cold surface, damage from cutting or stabbing, or multiple points being touched simultaneously, thanks to its ability to detect signals from over 860,000 tiny pathways in the material.
The researchers utilized a mix of physical testing and machine learning approaches to help the robotic skin ‘learn’ which routes are most important, allowing it to detect different types of touch more effectively.
In addition to possible future uses for humanoid robots or human prosthetics requiring a touch sensation, the researchers believe the artificial skin might be valuable in sectors ranging from automotive to disaster assistance.
Electronic skins transform physical information, such as temperature or pressure, into electrical signals. Multiple types of sensors are usually required for different types of touch; for example, one kind of sensor is required to detect temperature, another for pressure, and so on.
These sensors are then incorporated into soft, flexible materials. However, the materials are readily broken, and the signals from these many sensors might interfere with one another.
Having different sensors for different types of touch leads to materials that are complex to make. We wanted to develop a solution that can detect multiple types of touch at once, but in a single material.
Dr. David Hardman, Department of Engineering, University of Cambridge
“At the same time, we need something that’s cheap and durable, so that it’s suitable for widespread use,” added study co-author Dr. Thomas George Thuruthel from UCL
Their technique makes advantage of multi-modal sensing, a single type of sensor that responds differently to various forms of touch. Although determining the reason for each signal is difficult, multi-modal sensing materials are more reliable and simpler to manufacture.
The researchers melted down a soft, stretchy, and electrically conductive gelatine-based hydrogel and molded it into the shape of a human hand.
They experimented with various electrode configurations to find out which provided them with the most valuable information regarding various touch types. Due to the conductive material's tiny passageways, scientists were able to gather nearly 1.7 million bits of information from just 32 electrodes implanted at the wrist.
The researchers then examined the skin using different kinds of touch, including blasting it with a heat gun, pressing it with their fingers and a robotic arm, softly touching it with their fingertips, and even ripping it apart with a knife. To help the hand understand the meaning of the various kinds of touch, the researchers trained a machine learning model using the data collected during these experiments.
“We’re able to squeeze a lot of information from these materials – they can take thousands of measurements very quickly. They’re measuring lots of different things at once, over a large surface area,” added Hardman.
“We’re not quite at the level where the robotic skin is as good as human skin, but we think it’s better than anything else out there at the moment. Our method is flexible and easier to build than traditional sensors, and we’re able to calibrate it using human touch for a range of tasks,” said Thuruthel.
The researchers intend to do more testing on actual robotic tasks and enhance the electronic skin’s resilience in the future.
The Royal Society, the Engineering and Physical Sciences Research Council (EPSRC), a division of UK Research and Innovation (UKRI), and the Samsung Global Research Outreach Program provided funding for the study. Fumiya Iida attends Cambridge's Corpus Christi College as a fellow.
Journal Reference:
Hardman, D., et al. (2025) Multimodal information structuring with single-layer soft skins and high-density electrical impedance tomography. Science Robotics. doi.org/10.1126/scirobotics.adq2303