This post was originally published on here
The robot is hard to see without a microscope. It’s small enough to rest on the ridge of a fingerprint and can operate in liquid for months. Inside this speck is a functioning computer, a sensor, and a motor—integrated into a single machine that can sense its surroundings and change its behavior on its own.
That combination, long pursued in the field of robotics, is described in two new peer-reviewed studies published in Science Robotics and Proceedings of the National Academy of Sciences. Together, the papers report what researchers say are the world’s smallest fully programmable, autonomous robots—machines that operate at roughly the same scale as bacteria.
“We’ve made autonomous robots 10,000 times smaller,” said Marc Miskin, an assistant professor at the University of Pennsylvania and senior author on both studies, in a press release. “That opens up an entirely new scale for programmable robots.”
A Complicated Tradeoff
The difficulty wasn’t just making things small. The problem is that physics gets weird when you shrink down. Shrinking power supplies, fragile mechanisms, and the laws of fluid dynamics all collide at the micro-scale. For a robot this size, water doesn’t behave like water; it acts more like thick syrup. Tiny propellers can stall out and hinges can snap. Legs that work beautifully on a human scale become fragile and useless.
“If you’re small enough, pushing on water is like pushing through tar,” Miskin said.
Because of this, previous microrobots relied on external controllers. They were pushed around by magnetic fields, acoustic waves, or chemical reactions. They could move, sure, but they couldn’t decide. Their “intelligence” was stuck in bulky equipment outside the petri dish.
The new work takes a different approach. The researchers redesigned both motion and computing from the ground up to match the physical limits of the microscopic world. One key change was abandoning moving parts altogether.
Rather than paddling through fluid, the robots generate an electric field. That field pushes ions in the surrounding liquid, which in turn drag nearby water molecules. The resulting flow carries the robot forward. The robot, in effect, moves by stirring its environment.
“It’s as if the robot is in a moving river,” Miskin said, “but the robot is also causing the river to move.”
The Brain of Operation
Movement is physics, but autonomy is computation. For the robot to think, it needed a brain small enough to fit on board.
That brain came from the University of Michigan, where engineers have spent years shrinking computers to the absolute limit. The processor inside each robot runs on about 75 nanowatts of power—roughly one hundred million times less than a typical light bulb.
“We saw that Penn Engineering’s propulsion system and our tiny computers were just made for each other,” said David Blaauw, a senior author on the studies, in a press release.
To operate on such little energy, the team had to redesign computing itself. Conventional instructions were compressed into specialized commands like “sense temperature” or “move for N cycles.” What would normally take dozens of steps was folded into one.
“We had to totally rethink the computer program instructions,” Blaauw said. “Condensing what conventionally would require many instructions for propulsion control into a single, special instruction.”
Powering this tiny brain is light. Microscopic solar cells cover most of the robot’s surface. Pulses of light both energize the robot and program it. Remarkably, each robot carries a unique address, meaning researchers can send specific instructions to specific machines, even if they are swimming in the same dish.
This architecture allows the robots to operate independently for months. This includes executing digitally defined algorithms, storing short programs in memory, and updating their behavior based on sensor data.
Improvise, Adapt, Overcome
To prove these specs were truly autonomous, the researchers gave them a sense familiar to biology: temperature.
Cells generate heat as they metabolize or react to stress, so temperature is a vital metric in the microscopic world. In experiments, the robots were programmed to measure temperature, store the data, and report it back.
The robot then relays this information through a little dance.
“To report out their temperature measurements, we designed a special computer instruction that encodes a value in the wiggles of a little dance the robot performs,” Blaauw said.
The robots were also programmed to adapt. When placed in a fluid with a temperature gradient, they changed their motion depending on whether they detected warmer or cooler conditions, effectively climbing the gradient without external control.
A Milestone, With Limits
For years, researchers have built tiny machines that swim, crawl, or spin. Many are elegant, but most lack the “trinity” of robotics: true autonomy, reprogrammability, and onboard sensing.
These new robots aren’t perfect. They aren’t as fast as some chemical micromotors, they can’t yet handle complex environments like blood, and their memory is limited. But experts see this as a massive threshold crossed. By integrating propulsion, sensing, and computation into a single sub-millimeter device, the team has proven it can be done.
“This is really just the first chapter,” Miskin said. “We’ve shown that you can put a brain, a sensor, and a motor into something almost too small to see, and have it survive and work for months.”
