Swarms of microrobots show there is power in numbers

Ever since Richard Feynman’s famous 1959 speech, “There’s Plenty of Room at the Bottom,” scientists have considered what it would take to build a swallowable surgeon—one small enough to travel through our blood vessels and controlled enough to perform procedures at precise spots in the body.

They’ve made progress toward Feynman’s vision, to be sure: During the early 2000s, researchers designed individual catalytic micromotors that could propel themselves through liquid by creating gas bubbles from chemical fuels. And by 2009, a team at the Swiss Federal Institute of Technology (ETH), Zurich, had created a helical bot that was recognized by Guinness World Records as the “most advanced mini robot for medical use.” The tiny swimmer, about 20 μm long, made from semiconductor materials and controlled by a magnetic field, was designed to mimic the swirling flagella that bacteria use to move around (Appl. Phys. Lett. 2009, DOI: 10.1063/1.3079655).

Still, there’s a big difference between being the most advanced bot of the day and performing surgery as Feynman envisioned. Scientists are realizing that even the most sophisticated single swallowable surgeon won’t be sufficient to achieve that goal. Instead, they’re coming around to the idea that there is power in numbers.

As Metin Sitti, a microrobot expert at the Max Planck Institute for Intelligent Systems, sees it: “Swarming is indispensable for the translation of microrobots into the clinic.”

Just as ants in a colony can link their bodies to build a bridge across a gap, swarms of tiny bots can do things that individual bots can’t, researchers say. For instance, together, microrobots in a large group could have a higher propulsion speed than lone bots trying to fight their way through the viscous fluids inside us. Swarms of microrobots would also emit a higher signalthan individual bots, enabling scientists to track them in the body.

Li Zhang and Lixin Dong are two devotees of the swarm. They were there a decade ago in Bradley Nelson’s lab at ETH Zurich when the Guinness World Records bot was created. Since then, Zhang has gone on to lead a microrobot lab of his own at the Chinese University of Hong Kong. And Dong recently moved nearby, joining the City University of Hong Kong.

The duo understands that there are both benefits and drawbacks to the tiny size of microrobots. Sure, being microscopic means being able to fit inside a person’s blood vessel. But it also means sacrificing the inertial force that helps larger objects move through a liquid. Inertia keeps objects in motion. Because microrobots have a large surface area compared with their ultrasmall volume, the inertial force becomes negligible, and the external magnetic field propelling them forward has to fight against the drag caused by their large surface force.

This power struggle means that, to a single microrobot moving through a nonviscous liquid like water, it would feel like trying to move through honey or molasses.

“That’s why the first generation of our robots was designed as helical,” Dong explains. Bots with a corkscrew-like shape can in theory “drill” through liquid, he says.

Changing a microrobot’s shape is one way to facilitate its movement. Another way is to build magnetic robots and guide them through liquids with a magnetic field. If you’re talking about the liquid inside a human body, though, there’s a limitation to the strength of the magnetic field that can be applied safely.

Swarms of microrobots, on the other hand, are made of a large number of individual bots that move cohesively. Hypothetically, they could gain speed when moving through liquid if individual particles use one another as a counterforce. Currently, Sitti says, individual microrobots can swim through liquids at about 1 μm to 1 mm per second. To be able to fight against the slowest flow of blood in the human body, that speed must ramp up to at least 200 μm/s, he says.

While Zhang agrees that making microrobots move faster is an important step in translating them into humans, he’s also focused on getting them to where they need to go in the body. With swarms, this involves being able to manipulate the shape of the overall swarm so it can fit through blood vessels and into other areas.

Last year, Zhang and colleagues manipulated a swarm of magnetite (Fe₃O₄) nanoparticles, each about 500 nm in diameter, to change its shape from an amorphous blob of swirling bots to a thin ribbon, in which all the bots assembled end to end in longitudinal lines. Using an oscillating magnetic field, the team then split the ribbon of particles into subswarms and guided them each into parallel fluidic channels. After passing through the channels, the swarms reunited on the other side, assembling back into one large group (Nat. Commun. 2018, DOI: 10.1038/s41467-018-05749-6).

Jiangfan Yu, a postdoctoral researcher in Zhang’s group, says the team was able to split the swarm apart with help from fluidic drag. And the researchers could use the attractive magnetic forces between the individual bots to merge the subswarms after they passed through the channels.

Another potential benefit that researchers see for microrobot swarms is the ability to deliver larger payloads of medicine in the body. As a proof of concept, Dong and two researchers from Harbin Institute of Technology, roboticist Hui Xie and chemist Qiang He, directed a swarm of peanut-shaped microrobots made of hematite (Fe₂O₃) to form a vortex. Driven by a rotating magnetic field, the vortex swarm pushed a polystyrene sphere through liquid. The sphere was about 40,000 times the volume of an individual 3 μm long, 2 μm wide microrobot (Sci. Robot. 2019, DOI: 10.1126/scirobotics.aav8006). Although the researchers didn’t directly measure the speed at which the sphere moved or how far it traveled, Dong says it could go faster or slower depending on the magnetic field’s frequency, and there’s no limit to how far it could travel.

The key to forming the swarm is the individual particles’ peanut-like shape. That shape, combined with various magnetic-field configurations, can transform the swarm from a swirling vortex to a forward-marching lateral chain to a sweeping longitudinal ribbon. And this multimode formation gives rise to multimode function. Vortices can transport big targets, chains can pass through narrow gaps, and ribbons can sweep small targets forward, paving the way for diverse operations in the human body, the researchers say.

To guide swarms of microrobots through the body and direct them to perform various operations, scientists need to be able to “see” them. Imaging individual microrobots is a particular challenge because of their low signal. Although there are groups developing higher-resolution imaging tools that would allow researchers to image a single microrobot, many agree that swarms are the more cost-effective solution. Another plus is that the magnetic fields typically used to guide swarms have high precision and can be integrated into existing instruments, such as magnetic resonance imaging (MRI) machines.

For example, Zhang and coworkers demonstrated that the increased signal from a swarm of microrobots makes it possible to track these agents with MRI in the stomachs of rats (Sci. Robot. 2017, DOI: 10.1126/scirobotics.aaq1155). They used microalgae dip coated in magnetite to create their biohybrid microrobots. In theory, the bots might one day function as drug-delivery vehicles. As organic matter, microalgae have features, such as biodegradation and biocompatibility, that are favorable for in-body applications, and the magnetite coating allows the biohybrid bots to respond to magnetic-field control.

Experiments in fibroblast cells from mice showed that microalgae bots with thin magnetite coatings degrade and leave behind small fragments of the iron oxide. They also exhibited low toxicity toward the cells: more than 80% of the cells lived normally after the researchers cocultured them with the bots for 48 h.

Researchers like Dong and Sitti agree that near-term applications of microrobots would be in no-flow or low-flow regions in humans, areas like the gastrointestinal tract or inside the eye. In fact, Sitti predicts that swarms for the GI tract could land in clinical trials within 10–15 years.

Some scientists have been making progress toward this low-flow goal. For example, Zhiguang Wu, now a chemist at Harbin Institute of Technology, and Peer Fischer, a physical chemist at the Max Planck Institute for Intelligent Systems, created a slippery Teflon-like coating on the surface of a helical magnetic microrobot to ease its passage into the eyeball (Sci. Adv. 2018, DOI: 10.1126/sciadv.aat4388). The eye’s vitreous body acts like a dense jelly, barring anything from getting inside to the retina, making targeted drug delivery challenging. Using pig eyeballs taken from a slaughterhouse, however, Fischer, Wu, and colleagues moved a swarm of their slippery microrobots, 10,000 strong, from the center of the eye to the retina within 30 min.

Xiaohui Yan of Xiamen University is also targeting the eye with his newly developed nanorobots, collaborating with ophthalmologists to potentially treat diseases such as age-related macular degeneration and diabetic retinopathy. Current methods for treating these diseases involve injections or eye drops and deliver drugs by diffusion, so they aren’t efficient. Yan says his nanobots have shown excellent movement in the eye in preliminary results, as well as degradability and targeting capabilities, but he would not disclose further details about the bots themselves at this time.

Yan is upbeat about micro- and nanorobots moving into humans someday. These types of tiny robots can reach anywhere inside the human body because of their size, he says. Given time, he thinks, nano- and microrobots could function together to become the swallowable surgeon Feynman hoped for.

Cici Zhang is a freelance science writer based in China.