Not Your Grandma’s Robot

Professor’s insect-based design could revolutionize the world of robotics

Imagine a robot so small and light it could hover in the air on a fly’s wings. Robot expert Robert J. Wood thinks he can make it happen—and someday create enough of these to embark on rescue missions to save lives.

Inspired by the elegant aerodynamics of insects, Wood, assistant professor of electrical engineering, is working on creating small flying robots that he hopes will someday serve in search and rescue missions.

The faux insects, named Micro Air Vehicles (MAVs), even physically resemble insects, weighing less than a paperclip with a two- to three-centimeter wingspan. The only thing is, these micro-creatures haven’t yet left the ground.

Wood’s lab hosts a wide array of technology, from miniature helicopters for studying hovering mechanisms to high-power laser machinery for precisely cutting materials. The whiteboards overflow with scribbled equations and drawings.

While the thought of smart robotic insects may seem frightening, this is not Steven Spielberg’s “AI.” Wood wants these robots to able to perform tasks on their own, perhaps delivering information from one point to another. Moreover, the MAVs’ small size limits the processing power that the devices can carry.

“Right now we’re developing decision-making procedures predicated on really simple control behavior—for example, avoiding obstacles or flying towards sensory stimuli,” says Wood, who is teaching Engineering Sciences 159, “Introduction to Robotics.”

While even Wood concedes that his work may seem “science-fiction-esque,” there are wide-ranging applications for these flying machines.

For example, during a search and rescue mission for a nuclear disaster, where humans rescuers would be put in danger, hundreds of these MAVs could be sent into the area to search for survivors. With the ability to maneuver into small spaces within the debris, MAVs would be able to detect carbon dioxide gas and report the location of the survivors.

“Eventually we want the MAVs to be so inexpensive that they would be disposable. Even if most of them are destroyed in the search process, the mission would still be considered a success,” Wood says.

BRAINSTORMING BUGS

“What’s most impressive is the range of expertise he brings to the creation of these micro-creatures—he uses structural design, fluid mechanics, automatic controls, microfabrication, and more,” McKay Professor of Engineering Robert D. Howe writes in an e-mail. “By drawing on all these disciplines, he can create sophisticated devices with astonishing performance.”

The young and bearded Wood, who joined the Harvard faculty as an assistant professor in 2005, brought the idea for the project from his graduate work at University of California at Berkeley. The project’s initial goal was to create a small device capable of sustained autonomous flight.

“Flying insects have been doing it for thousands of years, so it seemed like an ideal place to start our design,” Wood says.

Wood’s research is highly interdisciplinary. By interacting with biologists, he has been able to study the mechanics of insect flight and mimic these biological processes using microrobotics technology.

“We’ve been using insights from biology as shortcuts to engineering,” he says.

While insects travel by varying modes of locomotion, Wood focuses on hovering, which he believes is the hardest, but most useful, insect behavior to mimic.

“Hovering is the most attractive feature for applications,” Wood says. “Surveillance, for example, requires the machine to hover and stare.”

However, even the slightest asymmetry in these machines can create huge body torques that would ruin the stability of device. While gears and switches work well when designing movement on a large scale, developing such intricate machinery is impossible at the microscopic level. To make flying and hovering possible in MAVs, Wood has developed special materials that offer varying flexibility.

“It’s kind of like a micro-origami process,” Wood says. “By altering the geometry of layers of different materials, we can produce different joint structures.”

HOVERING OVER THE GOAL

While the development of the flying machines is Wood’s main priority, he is also studying other methods of locomotion with the help of undergraduates.

Kyla J. Grigg ’07, who is working on developing a microrobotic fish, was initially attracted to Wood’s lab by his passion and excitement for robotics.

“He’s always very eager to help us with any problems we’re encountering, but he also allows us [undergraduates] a great deal of freedom to learn and explore the field ourselves,” Grigg says.

Jeffrey K. Ma ’07, who is developing algorithms to automate the hovering process, believes that an exciting aspect of the project is their ability to program the devices with senses not normally found in nature. Ma thinks it will be possible to make these robots even more sensitive than their organic counterparts.

Wood says that most aspects of the MAVs are functioning well independently, the next step in the project is to integrate the hovering processes, decision-making procedures, and external power.

“It may seem like a small task, but we’re still quite far from a finished device,” Wood says.

He expects that it will be at least a year before a prototype is complete—and his own insects finally take flight.

—Staff writer Anupriya Singhal can be reached at asinghal@fas.harvard.edu.