Robotic Touch: New Sensor Brings Human-Like Sensitivity to Machines

As robots become more integrated into daily life, the need for sophisticated tactile sensing – a ‘sense of touch’ – grows increasingly critical. Researchers at the University of Cambridge have developed a miniature tactile sensor that mimics the sensitivity and complexity of human fingertips, potentially revolutionizing robotic manipulation and beyond. The technology, detailed in Nature Materials, combines liquid metal composites and graphene to create a ‘skin’ capable of detecting not only force but also direction, slip, and surface texture.

The Challenge of Robotic Touch

Even as robots excel in vision and movement, replicating the human sense of touch has proven remarkably difficult. Human fingers rely on a network of mechanoreceptors to perceive pressure, force, vibration, and texture simultaneously. Reproducing this multidimensional perception in artificial systems requires devices that are small, durable, and capable of distinguishing between different types of forces. “Most existing tactile sensors are either too bulky, too fragile, too complex to manufacture or unable to accurately distinguish between normal and tangential forces,” explains Professor Tawfique Hasan from the Cambridge Graphene Centre, who led the research.

A Bio-Inspired Design

The Cambridge team’s innovation lies in a soft, flexible composite material. This material combines graphene sheets, deformable metal microdroplets, and nickel particles embedded in a silicone matrix. Inspired by the microstructures found in human skin, the researchers created tiny pyramid shapes, some as small as 200 micrometers across. These pyramids concentrate stress at their tips, allowing the sensor to detect extremely small forces while maintaining a wide measurement range. The resulting sensor is sensitive enough to detect a single grain of sand.

Key Capabilities of the New Sensor

• High Sensitivity: Detects forces at a scale comparable to human fingertips.
• 3D Force Vector Reconstruction: Can mathematically reconstruct the full three-dimensional force vector in real time by measuring signals from four electrodes beneath each pyramid.
• Slip Detection: Distinguishes between shear forces and normal pressure, enabling the detection of slipping objects.
• Versatility: Improves size and detection limits by roughly an order of magnitude compared to existing flexible tactile sensors.

Applications Across Industries

The potential applications for this technology are vast. In robotics, the sensor allows for more dexterous manipulation of objects, enabling robots to grasp fragile items without crushing them. Unlike conventional force sensors, this new system adapts in real time through slip detection. Beyond grasping, the sensor can identify the mass, geometry, and material density of tiny objects, opening doors for applications in microrobotics and minimally invasive surgery, where conventional sensors are too large.

The technology also holds promise for advancements in prosthetics. Highly sensitive, miniaturized 3D force sensors could enable more natural interactions with objects for prosthetic limb users, improving control, safety, and confidence.

Future Developments

Researchers are exploring ways to further miniaturize the sensors, potentially to below 50 micrometers, approaching the density of mechanoreceptors in human skin. Future iterations may also integrate temperature and humidity sensing, creating a fully multimodal artificial skin. As robots move into more complex and unpredictable environments, advancements in tactile sensing will be crucial for enabling truly intelligent and adaptable machines.

A patent application for the technology has been filed through Cambridge Enterprise. The research was supported by the Royal Society, the Henry Royce Institute, and the Advanced Research and Invention Agency (ARIA).

Reference

Yun, G., Chen, Z., Chen, Z., Chen, J., Zhou, B., Xiao, M., Stevens, M., Chhowalla, M., & Hasan, T. (2026). Multiscale-structured miniaturized 3D force sensors. Nature Materials.