The new technology, based on graphene and liquid metal composites, allows robots to detect not only how much pressure they exert on an object, but also the direction of the applied forces, whether an object is sliding, and even how rough a surface is. Most importantly, all of this is achieved on a scale small enough to rival the spatial resolution of human fingertips.
Flexible tactile sensors are essential for applications ranging from advanced prosthetics and wearable haptic devices to robots with highly dexterous hands. Human fingers use different types of mechanoreceptors to simultaneously perceive pressure, shear forces, vibrations, and texture. Replicating this level of multidimensional tactile perception in artificial systems has been a major challenge, especially in devices that need to be miniaturized and rugged for practical use.

Overcoming the limitations of current tactile sensors
: “Many existing multidimensional tactile sensors are too bulky, fragile, or complex to manufacture, or they cannot accurately distinguish between normal and tangential forces,” explained Professor Tawfique Hasan of the Cambridge Graphene Centre and lead author of the study. “This has been a major barrier to achieving truly precise robotic manipulation.”
To overcome these limitations, the team developed a new class of anisotropic porous elastomers synergized with graphene.

A graphene-liquid metal material inspired by human skin.
These soft, flexible composites combine graphene sheets, deformable microdroplets of liquid metal, and magnetically aligned nickel particles embedded in a porous silicon matrix.
Each component plays a key role:
Graphene creates flexible conductive pathways.
The liquid metal acts as a deformable conductive node.
The nickel particles, aligned by a magnetic field, provide directional sensitivity to the material.
The porous structure further enhances sensitivity by allowing the material to deform easily under minimal forces.
Drawing inspiration from the microstructures of human skin, the researchers molded the material into microscopic pyramids, some only 200 micrometers in size. These pyramids concentrate stress at their tips, enabling the sensor to detect extremely small forces while maintaining a wide and highly linear measurement range.

High sensitivity and true 3D force detection.
The result is a tactile sensor capable of reconstructing the complete force vector in three dimensions, enabling multidimensional tactile perception with exceptional performance:
Sensitivity up to 110 kPa⁻¹,
linear range of 500 kPa,
force direction error of less than 2 degrees, and
a detection limit of less than one micronewton—enough to detect a grain of sand.
Compared to existing flexible 3D force sensors, the new device improves both size and detection limits by approximately an order of magnitude.
Most importantly, the sensor can distinguish between shear and normal forces, allowing it to detect when an object begins to slip. By measuring the signals from four electrodes located beneath each pyramid, the system can mathematically reconstruct the complete three-dimensional force vector in real time.

From delicate grips to microrobots and prosthetics:
In practical demonstrations, the team integrated the sensors into robotic grippers. The robots were able to grasp fragile objects, such as very thin paper tubes, without crushing them. Unlike traditional sensors that require prior information about the objects, the new system adapts in real time by detecting slippage.
At even smaller scales, arrays of microsensors were able to identify the mass, geometry, and material density of tiny metal spheres by analyzing the 3D force distribution.

This opens the door to applications in:
Micromanipulation,
Minimally invasive surgery
, Microrobotics
where conventional force sensors are too large.

Impact on Robotics and Prosthetics of the Future:
The potential impact is enormous as robotics continues to expand in industry and society. According to the International Federation of Robotics, the global market for industrial robotics is valued at approximately £12.5 billion, and one of the key trends for 2026 is the increasing use of robots to address labor shortages.
Advances in tactile sensors like this could allow robots to interact with objects more safely and intelligently in real-world environments.
Beyond robotics, this technology could also transform the field of advanced prosthetics. Modern artificial limbs increasingly rely on tactile feedback to provide users with a sense of touch. Miniaturized, highly sensitive 3D sensors could enable more natural interactions with objects, improving control, safety, and user confidence.

“Our approach demonstrates that bulky mechanical structures and complex optical systems are not needed to achieve high-resolution 3D touch sensing,” explained Dr. Guolin Yun, the study’s first author. “By combining smart materials with skin-inspired structures, we achieved performance remarkably close to human touch.”
Looking ahead, the researchers believe the sensors could be miniaturized even further, possibly to below 50 micrometers, approaching the mechanoreceptor density of human skin. Future versions could also integrate temperature and humidity sensing, moving toward a fully multimodal artificial skin.
As robots move from controlled industrial environments into homes, hospitals, and unpredictable settings, advances in the sense of touch like this could be transformative, enabling machines not only to see and act, but also to feel.

References:

(Open access) https://www.nature.com/articles/s41563-026-02508-7

https://ifr.org/ifr-press-releases/news/top-5-global-robotics-trends-2026

Source: University of Cambridge