How a robot shaped like a dragonfly could help with environmental concerns

Want to feel like you’re at the best kids’ birthday party in the world, with a kiddie pool, the latest in mood-ring technology, and remote-control dragonflies that sometimes go crazy and start spinning around? Or would you rather do cutting-edge scientific research, combining hydrogels, balloon actuators, and independent robotic environmental response to temperature, pH, and pollution?

Ha ha, false dichotomy: You can do both, if you’re fortunate enough to work in the lab of Shyni Varghese, professor of biomedical engineering, mechanical engineering and materials science, and orthopaedic surgery.

It starts, really, with the kiddie pool. “They got me in trouble,” Varghese says, laughing, of Ph.D. student Vardhman Kumar and postdoc Un Kyung Ko, who were designing a floating robot and realized they needed a testing facility bigger than any of the containers in Varghese’s lab. “Duke immediately asked me, ‘Why are you buying a pool for the lab?’ ” Because once you’ve designed a soft robot shaped like a dragonfly—DraBot, they call it—it’s going to need to skim across water and attend to its business, and no sink or tank gives it the space for that.

DraBot is a soft robot; made of silicone, it looks like a simple drawing of a dragonfly—long, thin body; double pair of wings near the front. The dragonfly came about because Varghese, whose lab creates things like self-healing hydrogels that build or lose chemical bonds in response to changes in environmental pH, took a notion before she got on an airplane.

“She just sent an e-mail from the airport waiting lounge,” Kumar recalls. “‘Hey, I want to meet you as soon as I get back, I have this cool idea.’ That’s all that was in the e-mail. And so, she comes straight back from the airport, and she was like, ‘I want us to think and see if we can make a soft robot that can have sensors.’ ”

From the beginning it was a side project: The lab is busy with materials science and other funded research, but Kumar loved the idea. “I do a lot of photography of dragonflies,” he says, “which is one of the reasons it came about to do a dragonfly.”

The notion was multifaceted. Using the kind of materials the lab created (including those self-healing hydrogels, which form a strong bond at a certain pH and then release the bond if pH changes), Varghese wanted to create a robot that could independently interact with the environment and gather information. The group settled on a soft robot because soft robots can do more than your hard-corner metal guys. “They are squishy,” Kumar says. “So they can get through spaces that hard robots cannot.” They can squeeze their way into tight places. Moreover, “because they are lightweight, they can float much easier.”

DraBot is a couple of inches long, with an inch-and-a-half wingspan, and its two sets of wings are key. Because the robot is soft, it has no electronics, but it’s designed with tiny air channels in the front wings, pointing backwards. The tubes empty directly to the back wings, so the air can’t produce propulsion. When separate microchannels direct air into balloon actuators, the back wings lift, the original channels propel it, and DraBot moves forward. One wing pair is painted with that self-healing hydrogel, which bonds when the pH drops in an acidic environment. Then wings on that side bond. And DraBot, with propulsion from only one wing’s channels, goes in a circle, indicating it’s encountered acid. If the pH rises back to normal, the wings separate again and DraBot can go smoothly forward. Thermochromatic pigment on the wings makes them change color at high temperatures, and white microporous sponges attached to the wings repel water but absorb oil, becoming brown when they do.

Thus DraBot, in its environment, responds to pH, temperature, and pollution and signals its findings. “It can find applications in places such as where freshwater acidification is a concern,” Kumar says, from acid rain or mining discharges. “If we can have environmental probes like this hovering around in the area, they can report back as soon as they find leakage.” Swarms of them could eventually not just detect but even clean oil or chemical spills.

The study that produced DraBot is a proof-of-principle study, of course: It’s connected to its controllers by long thin air tubes. To be useful in the environment, soft robots like DraBot will need to be independent, raising the issue of power. “With soft robots,” Kumar notes, “people are trying chemical reactions, or incorporating living cells that generate energy,” so scientists have got a way to go. With hard robots, power is as simple as strapping on a battery pack.

On the other hand, a hard robot would just sink in a kiddie pool.