From motion to memory: Researchers create soft machines that amplify movement and remember touch

 

Conventional soft actuators are often limited by weak force, small displacement, and slow response. To overcome these limitations, researchers have developed a new mechanical system that can amplify motion and remember external triggers through the interaction between magnets and elastic membranes.

A joint research team led by Prof. Jeong-Yun Sun from Seoul National University's Department of Materials Science and Engineering and Prof. Ho-Young Kim from the Department of Mechanical Engineering has developed a new class of soft actuator based on elasto-magnetic instability, a mechanism that couples magnetic attraction with elastic restoring force.

The study was published in Nature Communications on January 10, 2026, under the title "Elasto-magnetic instabilities for amplified actuation and mechanical memory," with Seong-Yu Choi and Ji-Sung Park contributing equally as co-first authors.

Soft actuators are widely studied as artificial muscles because they can deform safely and flexibly. In nature, organisms such as the Venus flytrap, bladderwort, and pistol shrimp overcome the limits of soft structures by storing energy and rapidly releasing it through mechanical instability.

Inspired by this principle, the SNU research team designed a Coupled Elasto-Magnetic Vibration system, or C-EsMV, composed of permanent magnets, elastic membranes, and an electromagnet. When magnetic attraction and elastic tension are properly balanced, the system enters a bistable regime, allowing it to switch between two stable mechanical states. A small electrical input can then trigger a sudden transition, converting stored elastic energy into large and rapid motion.

Unlike conventional electromagnetic actuators, whose displacement typically increases gradually with input current, the C-EsMV system exhibits a nonlinear, stepwise response. Once triggered, the amplified motion can persist even after the input is reduced, owing to inertia-driven hysteresis. This hysteretic behavior also enables mechanical memory, in which external stimuli such as touch or a nearby magnet switch the actuator from a standby state into a memorized state. Depending on the input condition, the memory can be either volatile or non-volatile.

Key features of the EsMV system

Motion amplification through elasto-magnetic instability (EsMI)

The central achievement of this study is the realization of EsMI, created by balancing two competing forces: the attraction between paired magnets and the restoring force of stretched elastic membranes.

In the coupled system, elastic energy is stored as the magnets are pulled together and then rapidly released once an energy barrier is overcome, producing a slingshot-like motion. This mechanism enabled large-amplitude vibration under the same electrical input, with kinetic energy conversion enhanced by more than three orders of magnitude compared with the non-coupled system and an energy conversion efficiency ratio reaching up to 700-fold under optimized conditions.

Amplified force and displacement in a soft and compact actuator

The amplified vibration can also be translated into useful mechanical work. In one experiment, a hammer-like indenter attached to the vibrating magnet struck a lightweight ball. Under identical input conditions, the coupled system transferred substantially more energy to the ball than the non-coupled system, increasing the ball's maximum potential energy by approximately 50 times.

The actuator also generated a sharp increase in impact force after passing the activation threshold. Near resonance, the impact force was further amplified and became strong enough to break a thin glass wall, showing that soft mechanical systems can produce strong and rapid motion when instability and inertia are deliberately designed into the system.

Programmable and energy-efficient design

The C-EsMV system provides a programmable route to energy-efficient actuation by using its instability threshold and input waveform rather than relying only on continuously increasing current.

The researchers found that concave-shaped waveforms, such as pseudo-Gaussian waveforms, reduced the input energy required for amplification and improved energy conversion efficiency by up to 64.4 times compared with a standard sine wave. This makes the mechanism promising for soft actuators operating in space- and power-constrained environments.

Mechanical memory without electronic circuits or software control

Beyond motion amplification, the C-EsMV system can physically store information about external stimuli. A brief mechanical or magnetic trigger can push the system from a weakened vibration state into an amplified vibration state, which can remain even after the trigger is removed.

The system demonstrated both volatile memory, in which the memorized state gradually decays, and non-volatile memory, in which a short trigger induces persistent amplified vibration until the system is manually reset. The team also built a 3 × 3 mechanical memory array, where individual cells could record where and when a stimulus had been applied.

Significance of the research

This study presents a new design strategy for soft actuators by treating mechanical instability and inertia as functional design elements rather than as undesirable side effects.

By engineering the balance between magnetic attraction and elastic tension, the C-EsMV system produces large, discrete responses without relying solely on stronger electrical input. Because the mechanism is based on force balance rather than a specific material or geometry, it could be extended to various soft robotic and adaptive mechanical systems.

The research team noted that this approach could be particularly suited for applications requiring discrete, energy-efficient responses to small stimuli with programmable thresholds, such as mechanical transistors, spike-signal processors, or memory-integrated devices.