Reviewed by Lexie CornerFeb 24 2025
Researchers at the University of California, Santa Barbara (UCSB) and Dresden University of Technology's Cluster of Excellence Physics of Life (PoL) have developed sets of robots that mimic living systems by functioning as smart materials with adjustable strength and shape, according to a study published in Science.
An artist's depiction of a material-like collective of robots forming a rigid device. Image Credit: Brian Long, University of California, Santa Barbara (UCSB)
We’ve figured out a way for robots to behave more like a material.
Matthew Devlin, Study Lead Author and Former Doctoral Researcher, Physics of Life, University of California, Santa Barbara
The collective consists of separate autonomous robots shaped like disks, resembling small hockey pucks. These robots are designed to assemble themselves into various shapes and exhibit adjustable material strengths.
The study team focused on developing a robotic material that could be both strong and rigid while also possessing the ability to flow when a different shape is needed.
Robotic materials should be able to take a shape and hold it, but also able to selectively flow themselves into a new shape.
Elliot Hawkes, Associate Professor, Department of Mechanical Engineering, University of California, Santa Barbara
However, until now, it has been impossible to reorganize the group to flow and change shape at will while the robots were tightly bound together.
The researchers were inspired by earlier studies on the physical structure of embryos, conducted by Otger Campàs, a former professor at UCSB and now the director of PoL at Dresden University of Technology.
Living embryonic tissues are the ultimate smart materials. They have the ability to self-shape, self-heal and even control their material strength in space and time. While at UCSB, his laboratory discovered that embryos can melt like glass to shape themselves.
Otger Campàs, Director, Physics of Life, Dresden University of Technology
His study at UCSB found that embryos can shape themselves by transitioning between solid and fluid states, akin to melting glass.
He added, “To sculpt an embryo, cells in tissues can switch between fluid and solid states; a phenomenon known as rigidity transitions in physics.”
The ability of cells to organize during embryonic development transforms the organism from a mass of undifferentiated cells into distinct structures, such as hands and feet, and varied materials, like bones and brain.
The researchers focused on three biological processes underlying these rigidity transitions: the active forces that cells apply to each other to enable movement, the biochemical signaling that coordinates these movements in time and space, and the ability of cells to adhere to one another, contributing to the stiffness of the organism's final form.
Magnets integrated into the exterior of the robotic units simulate cell-cell adhesion in robotics. These magnets enable the group to function as a stiff material, with the robots adhering to one another. Eight motorized gears on each robot's exterior allow additional forces between robots, encoding these forces into tangential forces.
By varying these forces, the researchers were able to allow previously rigid robotic collectives to reconfigure and change shape. The introduction of dynamic inter-unit pressures helped overcome the challenge of making stiff robotic collectives more flexible, mimicking real embryonic tissue.
Biochemical signaling functions like a global coordinate system.
“Each cell “knows” its head and tail, so then it knows which way to squeeze and apply forces,” Hawkes explained.
This allows cells to alter the tissue’s shape, as demonstrated when cells align to lengthen the body. Light sensors with polarizing filters on each robot help the robots achieve this by determining which way to spin their gears and adjust shape when exposed to light.
“You can just tell them all at once under a constant light field which direction you want them to go, and they can all line up and do whatever they need to do,” Devlin added.
The researchers were able to manipulate the group of robots to behave like a smart material. Some parts of the group adhered to form a rigid material, while other parts activated dynamic forces to fluidize the collective.
The researchers developed robotic materials that can withstand large weights while reshaping, manipulating objects, and even self-healing by adjusting these properties among the robots.
The proof-of-concept robotic group currently consists of 20 relatively large units. However, simulations conducted by former postdoctoral associate Sangwoo Kim, now an assistant professor at EPFL, suggest the system could scale to include smaller units.
This could enable the development of robotic materials made of thousands of units that can assume various shapes and adjust their physical properties, offering new possibilities in the understanding of objects.
Combining these robotic ensembles with machine learning could lead to remarkable capabilities in robotic materials, fulfilling a science fiction vision, and extending to fields beyond robotics, such as active matter studies in physics or collective behavior in biology.
Journal Reference:
Devlin, R, M., et al. (2025) Material-like robotic collectives with spatiotemporal control of strength and shape. Science. doi/10.1126/science.ads7942