Kyla Hunter, engineering student, Duke University, and blog writer, Duke Research Blog
Imagine a robot small enough to fit on a US penny. Now, imagine a robot small enough to rest on the chest of Lincoln, not the Lincoln whose head decorates the front side of the penny but the even tinier version of him on the back of the penny, inside the Lincoln memorial.
Before it was changed to a Union Shield, the tail side of pennies featured the Lincoln Memorial, including a miniscule representation of the seated Lincoln statue that rests inside. Barely visible to the naked eye, this miniature Lincoln is on the order of a few hundred micrometres wide. As incredible as it sounds, this is the scale of robots being built by Professor Itai Cohen and his lab at Cornell University. On February 22, Cohen shared several of his lab’s cutting-edge technologies with an audience in Duke University’s Schiciano Auditorium.
To begin, Cohen described microscopic robots as consisting of two distinct parts: the brain of the robot and the brawn. The brain refers to the microchip, and the brawn refers to the legs or actuating limbs. Of these two, the brain, believe it or not, is the easy part. As Cohen explained, “Fifty years of Moore’s Law1 has solved this problem.” We now possess the ability to create ridiculously small microcircuits that fit the footprint of a few micrometres.
The brawn, on the other hand, is a major challenge. This is where Cohen and his lab come in. Their idea was to use standard fabrication tools used by the semiconductor industry to build the chip, and then build the robot around the chip by folding the robot into the 3D shape they desired. Think origami, but at the microscopic scale.
Like any good origami artist, Cohen and his collaborators recognised that it all starts with the paper. Using the unique tools at the Cornell Nanoscale Science and Technology Facility, they created the world’s thinnest papers, including one made of a single sheet of graphene. To clarify, that’s a single atom thickness.
Next, came the folding. As Cohen said, there’s really two main options. The first is to shrink the origami artist to the microscopic level but he conceded that science doesn’t know how to do that quite yet. Alas, the second is to have the paper fold itself. I will admit that as an uneducated listener, option number two sounds about as absurd as one. Regardless, it turns out to be a viable one.
The basic process works as follows. A 7 nm thick platinum layer is coated on one side with an inert material. When this is put in a solution and voltage is applied, ions that are dissociated in the solvent will absorb onto the platinum surface. This absorption creates a stress that bends the device. Reversing the voltage drives away the ions and unbends the device. Applying stiff elements to certain regions restricts the bending to occur only in desired locations. Devices about the thickness of a hair can be folded and unfolded using this method.
Origami microrobots in various configurations have been fabricated by Cohen and his collaborators.
As incredible as this is, there is still one defect, which is that these devices require a wire to an external power source that powers the folding. To solve this problem, Cohen and his team used photovoltaics (mini solar panels) that attach directly onto the device itself. When light shines on the photovoltaic (via sunlight or lasers), it moves the limb. With this advance, the team was able to develop BroBot, a robot that Cohen said, “flexes his muscles and looks like he belongs on a beach somewhere” when light shines on the front photovoltaics, then later OptoBot, the 2020 Guinness World Records-winning, smallest walking robot. OptoBot measures about 70 μm long depending on the design, 40 μm wide and 5 μm high, and can fold itself up and walk off the page.
1 of 2
BroBot, a robot “that flexes his muscles” when light is shined on the front photovoltaics.
2 of 2
OptoBot, the Guinness World Records-winning, smallest walking robot, measuring 40–70 x 40 x 5 μm.
With BroBot and OptoBot, Cohen and his team successfully eliminated the need for any external wire, but there was another important objective the team wanted to meet. These robots still required lasers to activate the limbs. In this sense, as Cohen explained, “They were still just marionettes being controlled by strings in the form of laser pulses.” To go beyond this, the team began working with X-Fab, a commercial foundry, to create microchips that would act as a brain that could coordinate the limb movements. In this way, the microscopic robots would be able to move on their own, without lasers pointed at specific photovoltaics. Cohen described this development as “Cutting the strings on the marionette and bringing Pinocchio to life.”
This was the final key step in the development of AntBot, a hexapod robot that moves all on its own. Cohen showed AntBot moving with a tripod gate, where it uses two legs from one side and a single leg from the other side to create alternating tripods that push the robot forward. All that has to be done is to place the microrobot in sunlight, and the brain takes care of the coordination.
AntBot, one of the most advanced robots to come out of the collaboration, can move autonomously, without the aid of lasers.
The potential for these kinds of microrobots is nearly limitless. As Cohen emphasised, the application for robots at the microscale is “basically anything you can imagine doing at the macroscale.” For example, cleaning surfaces, transporting cargo, building components and perhaps conducting microsurgeries or exploring new worlds that appear inaccessible.
One particularly promising application is mimicking the movement of cilia, the microscopic cellular hair responsible for countless locomotion and sensory functions in the body. A cilia-covered chip could become the basis of new portable diagnostic devices, enabling field testing that would be much easier, cheaper and more efficient2.
A schematic (a) and scanning electron microscope (SEM) image (b) of a surface covered by microscopic robotic cilia that deform back and forth under application of an alternating voltage to drive fluid across the surface.
Cohen and his collaborators envision a possible future where microscopic robots are used in swarms to clean blood vessels, help grow replacement tissues or probe large swathes of the human brain.
Duke Research Blog
References
1Moore’s Law was a prediction made by American engineer Gordon Moore in 1965 that the number of transistors per silicon (Si) chip would double every year. In 1975, as the rate of growth began to slow, Moore revised the timeframe to every two years. His revised prediction was two months out, with the number of transistors per chip having doubled circa every 18 months since 1961.
2Bolakhe, S. (2022). Cilia are minuscule wonders, and scientists are finally figuring out how to mimic them. Scientific American. July 11.Available at: bit.ly/40EjEEX