A team of engineers at Rice University and Kyung Hee University has developed a soft, shape-shifting mechanical surface that can respond to touch, sense its own movements and visually communicate changes in real time — an advance that could open new possibilities for human-machine interaction, wearable devices and immersive tactile displays.
Published in Science Advances, the research introduces a magnetically levitated mechanical metasurface capable of rapidly morphing into thousands of different shapes while remaining soft and durable enough to be pinched, pressed and twisted without damage.
The platform combines magnetic actuation, embedded sensing and LED-based visual feedback into a single programmable system designed for direct physical interaction between humans and machines.
“We spend a lot of time and resources improving computing power and rigid electronics,” said Raudel Avila, assistant professor of mechanical engineering at Rice and co-corresponding author of the study. “But humans are not rigid. We interact with the world through touch, deformation and movement. What we’re trying to do is engineer systems that communicate with people in a way that feels more natural and intuitive.”
Unlike conventional touchscreens and displays that primarily deliver visual information, the metasurface physically changes shape in response to interaction, creating a two-way flow of information through both touch and sight. Avila compared the experience to selecting an avocado at the grocery store.
“You don’t just look at an avocado to know if it’s ripe — you touch it,” Avila said. “That mechanical interaction gives you information. What excites us is creating surfaces that can also respond back and communicate their state visually and physically.”
The metasurface consists of a 6-by-6 array of soft elastomeric pixels controlled by electromagnets beneath the surface. By using attractive and repulsive magnetic forces, the system can raise or lower individual pixels with millimeter-scale precision, enabling more than 10 to the power of 30 possible surface configurations.
The researchers demonstrated the platform producing wavelike motions, ripple effects and checkerboard patterns as well as mimicking biological processes such as the rhythmic contraction and relaxation of the human heart. In another demonstration, the surface dynamically reshaped itself to guide droplets of water into different letters.
To enable real-time sensing, the team embedded inertial measurement unit sensors directly into the deformable surface. These sensors continuously monitor local surface tilt and reconstruct the overall shape of the metasurface without relying on external cameras or imaging systems.
The researchers also integrated a 7-by-7 RGB LED array that changes color depending on the surface’s shape and movement, effectively transforming the device into a dynamic three-dimensional display. In one demonstration, the system simulated ocean waves beneath a miniature paper boat while synchronized lighting effects visually represented moving water.
For Pei Liu, a doctoral student in Avila’s lab and lead author of the study, the project represents a new way of thinking about electronics and interfaces.
“We are not doing traditional electronics work,” Liu said. “We are introducing a new type of system that is soft and can deform into different shapes depending on what we want it to do.”
Liu likened the metasurface’s design to the human body, wherein the soft upper layer acts like skin, a more rigid lower layer provides structural support like bone, embedded magnets function like muscles and the sensors act like nerves that detect deformation and communicate information back to the system.
One of the team’s biggest engineering challenges involved precisely controlling the magnetic deformation of the surface in real time.
“When magnets get very close, the force becomes extremely strong, so controlling the movement is difficult,” Liu said. “We developed a simple analytical model that predicts how much voltage is needed to achieve a certain shape.”
That modeling framework dramatically accelerated the system’s responsiveness, reducing computation times from minutes to just seconds and enabling near real-time shape morphing.
The researchers say the technology could eventually support applications ranging from tactile educational tools and soft robotics to wearable systems, augmented and virtual reality interfaces and assistive technologies for people with visual impairments.
“We want to move beyond passive systems,” Avila said. “The goal is to create surfaces that don’t just display information but actively interact with people through both mechanical and visual feedback.”
This work was supported by the National Research Foundation of Korea, the Korea Research Institute of Chemical Technology, the Rice ENRICH Office and the Rice Space Institute.