These smart robots are smaller than a grain of salt

You wouldn’t know this teensy speck was a robot if you saw it. At less than a millimeter across, you might not notice it at all. But the itty-bitty machine offers big opportunities to explore the microscopic world. It’s the smallest robot that can move, think and act on its own, its creators say.

The mini robot was inspired by nature’s tiny, complex machines. “Cells and microorganisms are phenomenally sophisticated,” says Marc Miskin. “Nature has chosen this length scale to organize all of life.” An engineer, Miskin works at the University of Pennsylvania in Philadelphia. He hopes that similarly tiny robots will help uncover the secrets of the cellular realm.

The new robot is as small as a paramecium — a single-celled organism that lives in water. It isn’t the first machine less than a millimeter long, Miskin says. But it’s the first one that’s fully autonomous. Once programmed, it decides where to go, how to get there and what to do.

Such small robots might someday be able to travel through the human body to study cells or deliver drugs.

Designing the microbot was a challenge, says Miskin. Robots need sensors, computer processors and memory, power sources and pieces to control their motion. In large robots, these systems can be separated. But in tiny robots, everything is squished in close. And at such small scales, different forces affect how objects can move around. As a result, engineers had to rethink how to design this robot.

Robots reimagined

The new robot is powered by mini solar cells on its surface. From an LED overhead, it produces 100 nanowatts of power. Microwave ovens produce 10 billion times more! The tiny amount of power available to run the machine limited its design options. The team had to be clever about how the robot gets around and processes information, Miskin says.

To move, the robot needs to be in a liquid. It can’t swim, though. At very small scales, objects interact differently with liquids than we do. For a cell-sized thing, swimming through water feels more like swimming through tar. So the microbot uses a low-power process called electrokinetic propulsion. That means electricity (“electro”) drives its motion (“kinetic”).

Four electrodes on the robot can send and receive electric current. As current travels through the liquid from one electrode to another, it generates a force on any charged particles that are in the liquid. The force pushes the particles in one direction, but the liquid tries to pull them back. This creates a flow that carries the robot along with it.

To move in different directions, the robot sends current between different electrode combinations. It changes speed by adjusting the current strength.

Once their robot could move, it was time for the team to give it sensors and a “brain.” Temperature might be a useful thing for little robots to measure. That’s because it can be an indicator of cells’ health. To that end, the researchers attached a couple of tiny thermometers to their robot.

The microbot’s “brain” is a teeny computer with only a few hundred bits of memory. Most laptops, meanwhile, have at least 64 billion bits! Because of this, the designers had to write computer programs that used hardly any memory. They used tricks like organizing many small commands under one super command and writing programs on a normal desktop computer, then sending them to the robots.

Putting the robots to the test

In experiments, the cell-sized robot was programmed to measure the temperature of its surroundings and move from colder to warmer areas. When the liquid near the robot was cooled, the robot moved around until it found a warmer spot, just as programmed. When that area was cooled, the robot resumed its search for warmth.

Miskin’s group shared these results in Science Robotics last December. The microbots are also cheap and easy to make, the researchers note. They estimate that if made in large batches, the robots would cost just a penny each.

Hundreds of robots can be built at the same time for a low cost. Researchers can program the tiny machines individually or all at once.Maya Lassiter/University of Pennsylvania

Miskin’s team is now exploring whether these robots can safely travel inside the body. They’re also studying how the robots might talk to one another. “Cells by themselves are amazing,” Miskin says. But their real strength lies in working together, he notes.

These robots are a great development, says Veronika Magdanz. A biotechnologist at the University of Waterloo in Canada, she makes tiny medical robots. It’s a new field with many possibilities, she says. Tiny robots could someday break up blood clots and help diagnose diseases. She appreciates that these new designs can sense, move and compute on their own.

A big challenge with this tech is power, Magdanz adds. She points out that light-powered robots may not work well in the body. Still, she expects challenges like this to be solved. Sixty years ago, “nobody could think of making tiny robots smaller than what we can see,” she says. “Now we have the abilities and the technology to do that. That’s really cool.”