Scientists create metamaterials that learn shapes, adapt behavior, and move without central control like living systems.
Researchers at the University of Amsterdam have developed a new class of metamaterials that can learn to change shape, adapt their behavior, and even move without a central controller.Unlike conventional materials that respond in fixed ways to external forces, these engineered systems behave more like living matter.
They can adjust their responses based on past interactions, allowing them to perform tasks that typically require programmed robots.The materials are built as chains of identical motorized hinges connected by an elastic structure. Each hinge contains a microcontroller that tracks its own movement, stores past states, and communicates with neighboring units.This distributed setup allows the material to coordinate its behavior locally, rather than relying on a single control system. Over time, the structure learns how to respond to inputs by adjusting how each hinge moves and interacts.Materials learn and adaptThe metamaterials are trained through repeated interactions. Researchers bend certain hinges to act as inputs, while guiding others into a desired configuration. With each training cycle, the system updates how much force each hinge applies.Eventually, the material learns to reproduce the trained shape on its own whenever it senses the same input. It can also forget old configurations and learn new ones, or store multiple shapes and switch between them.“The most exciting observation of our research was that learning gives our metamaterials the ability to evolve – once the system starts to learn, the possibilities of where it ends up feel almost limitless,” said Yao Du.This ability to evolve behavior without centralized control marks a shift from programmable materials to systems that can adapt in real time.The system also highlights how intelligence can emerge from simple components working together, rather than relying on complex central processing. By distributing decision-making across the material itself, the researchers show how future machines could become more resilient, flexible, and capable of operating in unpredictable environments.Brainless systems gain intelligenceThe materials can perform tasks such as gripping objects or moving across surfaces, similar to simple living organisms. Earlier versions of such systems could move, but lacked the ability to learn or adapt.Each hinge applies torque based on local information, changing stiffness and preferred position. This allows the entire chain to reorganize itself dynamically in response to environmental conditions.Researchers say the next step is to move beyond static shapes and enable time-dependent behaviors, such as switching between movement patterns like crawling or rolling.“In future work, we aim to achieve learning time-dependent behaviour instead of changes into a static shape,” Du added.They also plan to explore how the system behaves under uncertainty, where learning occurs in noisy conditions and responses become probabilistic rather than fixed.The work reflects growing interest in materials that combine physical structure with adaptive intelligence, potentially opening new paths in robotics, soft machines, and responsive systems. The study is published in the journal Nature Physics.
Machine Learning Materials Science Metamaterials Nature Physics Robotics Smart Materials Soft Robotics
United States Latest News, United States Headlines
Similar News:You can also read news stories similar to this one that we have collected from other news sources.
How Wenger-Endorsed ‘Daylight Offside’ Rule, VAR Alternative Fared in First UsageA brand new season opened with new additions to officiating.
Read more »
Scientists Are Recycling Centuries-Old Bullets for a Badly Needed Toxic CompoundResearchers want to recycle more lead waste for high-tech solar panels via a new method.
Read more »
Regular Show Drops New Look at Cartoon Network RevivalRegular Show has dropped a new teaser for its new Cartoon Network revival, The Lost Tapes
Read more »
New studies of old dogs help scientists understand where they came fromScientists are tracing back the history of dogs using the oldest genes from the species ever studied
Read more »
Salesforce, Microsoft new products, pricing in new era for softwareBusiness Insider tells the global tech, finance, stock market, media, economy, lifestyle, real estate, AI and innovative stories you want to know.
Read more »
New pedestrian safety campaign begins in New York StateA new pedestrian safety campaign has begun in New York State.The $2.5 million campaign is led by the New York State Department of Health, in partnership with th
Read more »