Bioinspired Sensor Overcomes Sensitivity Trade-Offs

Reproducing aspects of the human sense of touch is vital for the development of everything from industrial robots to medical devices and virtual reality user interfaces. One of the most cost-effective ways to give machines the sense of touch is through the use of pressure sensors. But in the world of pressure sensors, one size does not fit all. In general, they can be tuned to detect very fine differences in pressure, or very large amounts of pressure, but not both. And that means that the machines that rely on these sensors struggle with versatility.

This situation undoubtedly has researchers in the field feeling the pressure. But a solution may finally be close at hand, thanks to the work of a group of engineers at Jilin University. The idea may not be entirely their own, as they borrowed heavily from nature, but their scorpion-inspired pressure sensor may be just what we need to give machines a more human-like sense of touch in the future.

Scorpions, despite their poor eyesight, have an extraordinary ability to perceive their surroundings. This is made possible through two specialized structures. First are trichobothria, long hair-like structures that respond to subtle changes in airflow, letting the animal feel predators or prey moving nearby. Second are slit sensilla, which are sensory neurons beneath tiny cracks in the exoskeleton that convert vibrations in the trichobothria into signals the nervous system can process. Together, these features allow scorpions to detect both delicate shifts in air currents and strong ground vibrations with impressive accuracy.

Inspired by these natural systems, the team designed a bioinspired piezoresistive pressure sensor (BPPS). The device addresses the classic engineering trade-off between sensitivity and range by combining two complementary elements: stress traps and flexure suppression units.

On the top side of a silicon chip, the researchers etched stress traps, which are microscopic structures that concentrate mechanical energy in much the same way the slits in a scorpion’s exoskeleton funnel vibrations. These traps greatly enhance the sensitivity of the sensor, allowing it to register even the faintest pressures. On the underside of the chip, they added flexure suppression units. These were modeled on the claw-like bases of scorpion hairs, which limit excess membrane bending and reduce mechanical noise. In the artificial version, the suppression units prevent distortions that normally reduce accuracy, thereby extending the linear range of pressures the device can measure.

The result is a sensor with both high sensitivity and excellent linearity across a wide span of pressures, from 0 to 500 kilopascals. In technical tests, it achieved a sensitivity of 65.56 millivolts per volt per kilopascal and maintained near-perfect linearity (a coefficient of 0.99934). It also demonstrated rapid response and recovery times, as well as durability over tens of thousands of cycles.

To move beyond the lab, the researchers integrated the BPPS into a six-legged robot and connected it to a deep learning network. Much like a real scorpion, the robot could sense minute air pressure changes around its body and respond accordingly. In demonstrations, it quickly moved away from large approaching objects, mimicking predator avoidance, and navigated toward smaller targets resembling prey.

By merging careful engineering with lessons from nature, the team has shown how scorpions’ ancient survival tools can guide the development of the next generation of high-performance sensors. If widely adopted, these devices could bring machines one step closer to experiencing the world as living creatures do.