Shape transfer robots can flow like liquid to support human body weight

In a major advance in robotics and materials science, researchers have developed a range of robots that can transition between solid and liquid states, thus supporting hundreds of times more weight than themselves when melted. This innovation brings us closer to programmable problems that can change their physical properties at will.

A team of researchers led by scientists at UC Santa Barbara and the University of Technology of Dresden drew inspiration from an unlikely source: developing embryos. Their findings, published in Science on February 21, demonstrate how simple robot units work together to create behaviors of similar materials seen only in science fiction.

“We have found a way to make robots more like materials,” said Matthew Devlin, a former doctoral researcher at Santa Barbara, the study’s lead author.

The system consists of disk-shaped robots, similar to small hockey pucks, which can quickly switch between sufficient rigidity to support large loads and enough fluid to flow into the new shape. This dual capability solves the fundamental challenge in robotics: creating structures that are both powerful and adaptable.

“Robot material should be able to shape and hold it up,” explains Elliot Hawkes, a professor of mechanical engineering at UC Santa Barbara.

Learn from nature

The breakthrough comes from studying embryonic development, where cells naturally demonstrate this ability to transition between liquid and solid states. “Living embryonic tissue is the ultimate smart material,” said Otger Campàs, director of life and life physics at Dresden Polytechnic University. “They have the ability to self-form, self-heal, and even control the material power of space and time.”

The researchers focused on three key biological processes: force cells used to move to each other, through the coordination of biochemical signals and their ability to join together. They then translate these natural mechanisms into engineering solutions.

Each robot in the collective has eight motorized gears around it and around the magnet to adhere to adjacent units. The force generated by gears between robots is similar to that between cells in living tissues, while magnets mirror cell adhesion. A light sensor with a polarization filter helps coordinate the movement of the robot, similar to how cells respond to biochemical signals during embryonic development.

From theory to practice

In the demonstration, the robots collectively showed significant abilities. It can form complex structures such as arches, repair gaps in their formation, and even act as tools by flowing around objects and then fixing them to manipulate them. Perhaps most impressive is that the system supports human weight (about 700 Newtons) before transitioning to a flow state, more than 500 times the weight of a single robot.

While the current prototype consists of only 20 relatively large units, computer simulations show that the system can be scaled to include thousands of miniaturized robots. This scaling potential may lead to truly programmable materials that can dynamically change their physical properties.

The research team’s work goes beyond robotics and provides insights into physics and biology. “The combination of these robots collectively with machine learning strategies that control them can generate significant capabilities in robotic materials, thus enabling the science fiction dream to come true,” said Sangwoo Kim, a former postdoctoral researcher in Campàs lab, now an assistant professor at EPFL.

This study was supported by the National Science Foundation of America and Germany’s German Excellent Life Physics Group (German Research Foundation).

This study represents an important step towards materials that can actively control their own characteristics, from construction equipment to medical devices that may revolutionize the field. As technology evolves, we may see structures that can adjust their shape and power as needed, opening up new possibilities in architecture, engineering and other aspects.