← Latest papers
💻 computer science

Pneumatic neurons for soft robots enable inflate-and-fire networks for rhythmic motion

This paper introduces "Pneu-rons," soft pneumatic modules that integrate logic, energy conversion, and actuation to form decentralized, electronics-free networks capable of generating adaptive rhythmic motion for soft robots through material-driven "inflate-and-fire" dynamics.

Original authors: Dongting Li, Michael Tolley, Nick Gravish

Published 2026-09-14
📖 7 min read🧠 Deep dive

Original authors: Dongting Li, Michael Tolley, Nick Gravish

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the natural world, movement often feels like a conversation between the body and the environment. A worm crawling through soil or an insect navigating a leaf does not rely on a single, central brain to issue every command. Instead, these creatures use distributed networks of nerves that run along their bodies, allowing different segments to react to local conditions while still moving in a coordinated rhythm. This biological strategy, known as a central pattern generator, allows for resilience and adaptability; if one part of the system is disturbed, the rest can often adjust without needing a top-down order. Soft robotics, a field dedicated to building machines from flexible, compliant materials rather than rigid metal and plastic, has long sought to replicate this kind of decentralized intelligence. However, most soft robots today still depend on bulky external pumps and complex electronic controllers to tell them when to move, which limits their ability to operate independently in remote or delicate environments.

Researchers at the University of California, San Diego have taken a significant step toward solving this problem by creating a soft robot that controls itself using the same principles found in nature, but without any digital electronics. They developed a new type of soft module they call a "pneumatic neuron," which functions as a self-contained unit capable of sensing, thinking, and acting all at once. By linking these modules together in a ring, the team created a system that generates its own rhythmic motion, much like the nervous system of a worm. The robot moves by inflating and deflating its segments in a traveling wave, driven entirely by the physics of heat and pressure rather than by a computer program. This approach allows the robot to adapt its movement to changes in temperature or physical load simply by reacting to its surroundings, offering a new path toward machines that are as robust and autonomous as the living creatures they mimic.

The core of this invention is a small, soft pouch filled with a special liquid that boils at a very low temperature. Inside each pouch sits a tiny heater and a mechanical switch made of flexible plastic. When electricity flows to the heater, the liquid inside the pouch turns into gas, causing the pouch to expand. This expansion is not just a movement; it is also a signal. As the pouch inflates, it presses against the internal mechanical switch. The switch is designed with two distinct thresholds: a lower point where it cuts power to the previous module, and a higher point where it sends power to the next module in the line. This creates a cycle the researchers call "inflate-and-fire." The module inflates, triggers its neighbor, and then cools down, allowing the gas to turn back into liquid so the cycle can begin again. Because the liquid boils at such a low temperature, the system can operate with very little energy, and because the switch is mechanical, it requires no microchips or software to function.

To test whether these individual modules could work together to create complex motion, the researchers connected them into a ring. They found that for the system to produce a stable, repeating wave of motion, the modules needed to perform three specific actions: they had to excite the next module, inhibit the previous one, and hold their own state until the next module was ready to take over. Without all three of these behaviors, the system would either get stuck in a single inflated state or the signal would die out before it could travel around the ring. When the team built a ring of just three modules, it successfully generated a continuous traveling wave. At any given moment, two of the three modules were active and warm, passing the signal along in a seamless loop. This "two-hot" state allowed the robot to sustain its rhythm indefinitely as long as it had a power source, proving that the network could self-organize without a central brain.

The true power of this system lies in its ability to adapt to the physical world. The researchers tested the robot by adding weight to specific modules and by changing the temperature of the environment. When they placed heavy weights on the robot, the added pressure made the mechanical switches trigger sooner, but the weight also made it harder for the modules to deflate. This changed the timing of the wave, slowing the robot down, but it did not stop it. The robot simply adjusted its pace to match the new load. Similarly, when they cooled the robot down in a freezer, the wave slowed and eventually stopped because the liquid could not boil fast enough. When they heated it on a hot plate, the modules inflated spontaneously, and the system entered a different state where it stopped oscillating and held a static shape. These experiments showed that the robot's behavior is not fixed by a pre-written code but emerges directly from the interaction between the material properties of the robot and its environment.

Perhaps the most striking demonstration of this system's resilience came when the researchers physically cut the ring in half. In a traditional robot, severing a control line would cause the entire machine to fail. Here, when the team cut the six-module ring into two separate three-module rings, both halves immediately began to oscillate on their own. One half continued its wave, and the other half, after a brief manual restart, also found its rhythm. The system did not need to be reprogrammed or rewired; the simple act of cutting the connection allowed the two new rings to form their own independent patterns. This suggests that these pneumatic neurons could be used to build robots that can survive damage, potentially splitting into smaller, functional units if they are torn apart, much like some animals that can regenerate lost limbs.

The researchers also attached small, L-shaped legs to the modules to see if the robot could actually move across a surface. As the modules inflated and deflated in a traveling wave, the legs gripped the ground when inflated and slid when deflated, creating a crawling motion similar to that of an earthworm. The robot moved forward by about 25 millimeters over six cycles, a speed determined by the time it takes for the liquid to boil and cool. While this is slow compared to a mechanical robot, the researchers noted that the speed is a property of the materials chosen, not the architecture itself. By changing the fluid or the heater, the timing could be adjusted without altering the fundamental logic of the system. This ability to generate locomotion without any external pumps or electronic controllers opens up possibilities for robots that can operate in places where carrying heavy equipment is impossible, such as deep underground or in space.

This work represents a shift in how soft robots are designed, moving away from the idea of a machine that is controlled by a separate computer toward a machine where the body itself is the computer. By embedding the logic of movement directly into the material and the physics of the actuator, the researchers have created a system that is inherently modular and robust. The robot does not need to "know" where it is or what the environment looks like; it simply reacts to the heat, pressure, and connections around it. While there are still challenges to overcome, such as finding materials that can withstand higher temperatures for faster movement, the pneumatic neuron offers a new blueprint for adaptive robotics. It suggests a future where machines can be as flexible and self-sufficient as the living organisms that inspired them, capable of navigating complex worlds through the simple, elegant physics of inflation and fire.

Drowning in papers in your field?

Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.

Try Digest →