Robot hand inspired by Venus flytrap senses objects before touching them

- KAIST Professor Moon Hong-chul's team develops material that bends and grips when exposed to light

- Single material combines sensing, actuation and shape retention for physical AI applications

Yang Hyun-woo (from left), Lee Kyung-rok, Kwon Jin-han, doctoral students, and Professor Moon Hong-chul. [Provided by KAIST]

South Korean researchers have developed a new soft-robot material modeled on the Venus flytrap — a plant that detects approaching prey and snaps its leaves shut in an instant.

KAIST announced Wednesday that a research team led by Moon Hong-chul, a professor in the Department of Chemical and Biomolecular Engineering, has developed an ionic soft-robot material called "Ionograsper" that can sense nearby objects and move in response to light.

When a person reaches for an object, the eyes and skin detect it before the muscles move. Robots work the same way, requiring sensors to perceive the environment and actuators to generate physical movement.

In soft robots, however, fitting both sensors and actuators onto a flexible, bendable body multiplies the number of components and wiring, making the overall structure more complex.

The research team solved this problem with a single material, embedding both the sensing and muscle functions into one soft polymer.

The key lies in combining azobenzene — a molecule that changes its structure when exposed to light — with a polymer that absorbs moisture from the air. Using these two substances, the team created a mesh-like structure through which positive and negative ions can move freely.

When a charged object approaches, the ions inside the material rearrange, generating an electrical signal. The principle is similar to what happens when a balloon rubbed against hair is brought close — the hair moves without being touched. This allows the material to detect an object's approach and movement even before contact.

Once an object is detected, shining ultraviolet light transforms the material into a gripping "robot hand." The light alters the molecular structure of the azobenzene, causing moisture to escape from one side of the material and making it bend. The team used this motion to successfully grasp real objects.

Notably, the material does not immediately return to its original shape when the light is turned off. Instead, moisture re-enters through nanoscale pores on the surface, and the polymer structure slowly relaxes, holding the deformed shape for more than 10 minutes. This means the material does not need continuous light exposure to maintain its grip, reducing energy consumption.

An illustration of the Ionograsper's operating principle and potential applications, inspired by the Venus flytrap (AI-generated image). [Provided by KAIST]

In effect, a single material performs all three functions — sensing, movement and shape retention. Because separate sensors and actuators are no longer needed, as they are in conventional soft robots, the number of components and wiring is reduced and the robot's structure is simplified.

The team plans to improve sensing range, actuation speed, durability under repeated use and gripping strength. It also aims to upgrade the technology so the material can be driven by visible or near-infrared light instead of the ultraviolet light currently used.

The researchers expect the material to find applications in soft grippers, wearable robots, human-machine interfaces and bioinspired robots — particularly if combined with physical AI systems that use AI to assess surroundings and translate that assessment into physical action.

"We designed a single material that senses objects and moves when exposed to light, without attaching separate skin and muscles to the robot," Moon said. "We will develop it further into a sensing and actuation material for physical AI robots by integrating AI control technology."

The findings were published in the international journal Advanced Materials.

nbgkoo@heraldcorp.com