Scientists at the University of Pennsylvania, working alongside researchers from the University of Michigan, have unveiled the world’s smallest fully programmable and autonomous robots, marking a major breakthrough in microscale robotics.
The tiny robots—described as micro-swimmers—measure just 0.2 x 0.3 x 0.05 millimetres, roughly the size of microorganisms. Despite their minuscule size, they can move independently, sense their surroundings, process information and respond to environmental changes, all while costing about one cent each.
The research, supported by the National Science Foundation, was published in Science Robotics and the Proceedings of the National Academy of Sciences (PNAS).
Powered by Light, Guided by a Micro Brain
The robots are powered entirely by light and run on an ultra-low-energy micro “brain” developed at the University of Michigan. Each robot contains solar panels that supply just 75 nanowatts of power—about 100,000 times less energy than a smartwatch uses.
Despite this extreme energy constraint, the robots can execute complex programs, sense factors such as temperature, and adjust their movement patterns accordingly.
“We’ve made autonomous robots 10,000 times smaller,” said Marc Miskin, assistant professor of electrical and systems engineering at Penn and senior author of the studies. “That opens up an entirely new scale for programmable robots.”
Swimming by Moving Water, Not Themselves
At such tiny scales, movement through water is incredibly difficult due to extreme drag—more like moving through tar than liquid. To overcome this, researchers designed a propulsion system with no moving parts.
Instead of pushing themselves forward, the robots generate an electrical field that nudges ions in the surrounding liquid. These ions then push water molecules, creating enough force to move the robot. This innovative mechanism, detailed in the PNAS study, makes the robots highly durable and capable of swimming for months at a time.
The robots can also move in complex paths and operate in coordinated groups, similar to schools of fish.
Individually Programmable, Infinitely Scalable
Each micro-robot carries a unique identifier, allowing scientists to program them individually using light pulses. This opens the door to deploying swarms of robots, each performing a specific role within a larger task.
To fit computing into such a tiny space, the Michigan team radically redesigned traditional programming instructions, compressing complex propulsion commands into single specialised instructions that fit within the robot’s microscopic memory.
“We had to totally rethink how a computer program works at this scale,” said David Blaauw, professor of electrical and computer engineering at the University of Michigan and senior author of the Science Robotics paper.
Potential Applications: Medicine to Manufacturing
Researchers say the technology could revolutionise medical diagnostics, enabling robots to monitor cell-level health, detect disease markers, or deliver treatments with extreme precision. Beyond medicine, the robots could help build tiny, highly accurate devices in advanced manufacturing and materials science.
“For decades, electronics have shrunk rapidly, but robots haven’t kept pace,” Miskin said. “This breakthrough finally closes that gap.”
What’s Next
The team now plans to explore how large groups of these micro-robots can work together on complex tasks—potentially reshaping the future of robotics at the smallest imaginable scale.