- Yale University discovers flowing liquids can remember past events, potentially changing engineering thinking.
- Microfluidic chip uses flexible fibers to record fluid's past states, no electronic components needed.
- The system can provide a new type of controller for soft robots, adapting to complex environments.
- Researchers hope to train liquid networks to produce specific responses, moving toward physical learning.
Yale University researchers have discovered a method that enables flowing liquids to remember past events, achieved by using tiny flexible fibers that change shape under pressure. The discovery could help engineers develop soft robots that operate without traditional electronic controllers. The concept upends conventional thinking about engineering problems. Most liquid systems react based on the conditions they encounter, but Yale's design can retain traces of previous flow. Its physical components effectively store information, allowing past events to influence the system's next behavior.
Inside the microfluidic chip, researchers placed flexible fibers within a network of narrow channels. Each channel is about half the depth of a human hair. As fluid moves through the channels, pressure pushes individual fibers between two stable configurations. Each position changes the path available for fluid to pass through, creating a physical record of the system's previous state. Engineers can represent these two configurations as binary digits — zero and one. Unlike conventional computer memory, the arrangement requires no electronic components to switch states. The study's lead author, Abhineet Rajput, a graduate student in the lab of Yale professor Amir Pahlavan, explained the distinction: "Traditional microfluidics relies on active components, while this system uses microfibers to passively store information."
How the microfluidic chip works
The fibers can also influence one another. Their interactions allow individual changes to shape the behavior of the larger network, producing more complex physical memory. The research may open another avenue for controlling soft robots. Such machines often need to bend, squeeze through tight spaces, or operate in environments that challenge conventional hardware. A fluidic controller can use pressure and flow rates to transmit signals. Channels connect different parts of the system, while flexible fibers regulate fluid motion and retain information. Pahlavan says the system holds particular value in environments where electronics face limitations.
"We can design a fully fluidic controller that can be integrated into a soft robot and effectively act as its brain," said Pahlavan, an assistant professor of mechanical engineering at Yale. Electromagnetic interference can disrupt electronic systems in certain operating environments, while a controller based on fluid motion offers an alternative, though researchers still need to prove it can reliably guide a robot in practical tasks.
Future research directions
Memory is only part of the researchers' ambition. The next question is whether engineers can train liquid networks to produce specific responses. Training can alter how the network reacts to incoming flow, enabling it to adapt its behavior. Achieving this would move the system beyond merely retaining past information toward physical learning. The same principles may also interest biologists studying how organisms without brains navigate their surroundings. Slime mold, for example, can find efficient paths through mazes. Researchers also hope to understand how fungi and plant roots respond to environmental changes.
Liquid networks may provide a physical model for studying how interactions and accumulated experience shape behavior. Whether the technology can ultimately learn useful tasks remains an open question.
The research results were published in the journal Physical Review X.
Item Specification Processor/SoC Microfluidic chip Memory Flexible fibers
Potential applications and challenges of liquid control systems
Yale University's research shows that technology using flowing liquids to record past states could provide new ideas for future soft robots. This liquid-based controller can demonstrate higher reliability when facing challenges such as electromagnetic interference. However, how to train these liquid networks to adapt to specific tasks remains a problem to be solved. This technology not only has potential impact in the engineering field, but may also attract the interest of biologists, because it provides a physical model for studying how organisms navigate without brains.