Skip to Content
Advertisement
Technology

The Wisdom of a Brainless Robot

How a soft-bodied machine uses physical intelligence to navigate.

3:36 PM CDT on September 11, 2023

Think of a pair of tango dancers, gliding across the floor, picking up subtle physical cues from one another. They move together effortlessly, like water, through a maze of carefully choreographed steps. The intelligence they possess is not just in their heads but in the muscles, tissues, and skin of the body. It’s physical.

Featured Video

Scientists are increasingly working to build this kind of physical intelligence into robots. For machines to achieve the kind of intelligence possessed by humans and other biological creatures, they ultimately need not just brain-like computational power, but also bodily smarts—the ability to pick up sensory cues from the environment and use that information to learn and make decisions.

We wanted to make a robot that could solve a maze by itself.

Recently, scientists from North Carolina State University achieved a breakthrough: They created a robot that is powered by physical intelligence alone. This “brainless” robot is the first that can navigate complex mazes without any need for computer or human guidance.

Advertisement

“We wanted to make a robot that could solve a maze by itself,” says Jie Yin, co-corresponding author of the paper and an associate professor of mechanical and aerospace engineering at North Carolina State University. Before now, robots could only navigate mazes within multi-walled spaces if they contained built-in electronics-based controllers and sensors, says Yin, or if they were manipulated by external sources of light or magnets. 

The robot Yin and his team designed does not look anything like the robot of popular imagination: It has a soft body, kind of like that of an octopus, made of ribbon-like liquid crystal elastomers and shaped like a piece of fusilli or rotini pasta. It moves by rolling, drawing energy from hot surfaces (131 degrees Fahrenheit at a minimum)—the hotter the surface, the faster it moves. Having no hard, rigid materials, edges, or structures allows it to navigate tight spaces and makes it safer for humans who want to interact with it. 

The researchers had already experimented with similar robots in a previous experiment, but these would get stuck in the maze between parallel objects. This time, they also made the design of the robot asymmetrical, which allows it to move in arcs and quickly get unstuck. “It can bend in any direction, change shape, and squeeze into a space that is smaller than the robot body itself,” says Yin. Think of a cup with a wide mouth and narrow base—it rolls along a circular rather than straight path.

Credit: Yin Lab@NCSU / YouTube
Advertisement

If this robot were outfitted with sensors and cameras, says Yin, it would be ideal for collecting data from a desert-like environment that is too hot for humans to navigate, or to assess the quality of roads, which also tend to store a lot of heat—even to complete missions on the moon, where temperatures can soar to almost 400 degrees Fahrenheit.

Yin’s robots may look more like pasta than people, but they possess a kind of intelligence that ChatGPT could only dream of.

Lead image: AlponaArt / Shutterstock

Advertisement

Stay in touch

Sign up for our free newsletter

Related Stories

Isaac Asimov Showed Us How to Avoid the AI Apocalypse

But the famed sci-fi author’s 80-year-old rules of robotics may need some updating

September 24, 2026

Can AI “Feel” Pain?

Simulated pain can make some AI models override instructions and put their own welfare ahead of humans

September 22, 2026

What Grok and Claude Have to Say about the AI Apocalypse

A maniacal villain and a polite librarian on the risk that AI could end humanity

September 16, 2026

A Process Made for Mars is Now Uncovering Ancient Art on Earth

And you can even run it on your own phone

September 15, 2026

Inside IBM’s New Quantum Computing Fridges, 180 Times Colder Than Deep Space

These bulky units can link qubits together like never before

September 11, 2026

AI Engineers Are Having Their Oppenheimer “Destroyer of Worlds” Moment

A cascade of regret and apocalyptic short-term forecasts hit social media this week

September 10, 2026