We take for granted the effortless flight of insects, thinking nothing of swatting a pesky fly and crushing its wings. But this insect is a model of complexity. After 12 years of work, researchers at the Harvard School of Engineering and Applied Sciences have succeeded in creating a fly-like robot. And in early May, they announced that their tiny RoboBee (yes, it’s called a RoboBee even though it’s based on the mechanics of a fly) took flight. In the future, that could mean big things for everything from disaster relief to colony collapse disorder.
The RoboBee isn’t the only miniature flying robot in existence, but the 80-milligram, quarter-sized robot is certainly one of the smallest. "The motivations are really thinking about this as a platform to drive a host of really challenging open questions and drive new technology and engineering," says Harvard professor Robert Wood, the engineering team lead for the project.
When Wood and his colleagues first set out to create a robotic fly, there were no off the shelf parts for them to use. "There were no motors small enough, no sensors that could fit on board. The microcontrollers, the microprocessors—everything had to be developed fresh," says Wood. As a result, the RoboBee project has led to numerous innovations, including vision sensors for the bot, high power density piezoelectric actuators (ceramic strips that expand and contract when exposed to an electrical field), and a new kind of rapid manufacturing that involves layering laser-cut materials that fold like a pop-up book. The actuators assist with the bot’s wing-flapping, while the vision sensors monitor the world in relation to the RoboBee.
"Manufacturing took us quite awhile. Then it was control, how do you design the thing so we can fly it around, and the next one is going to be power, how we develop and integrate power sources," says Wood. In a paper recently published by Science, the researchers describe the RoboBee’s power quandary: it can fly for just 20 seconds—and that’s while it’s tethered to a power source. "Batteries don’t exist at the size that we would want," explains Wood. The researchers explain further in the report: " If we implement on-board power with current technologies, we estimate no more than a few minutes of untethered, powered flight. Long duration power autonomy awaits advances in small, high-energy-density power sources."
The RoboBees don’t last a particularly long time—Wood says the flight time is "on the order of tens of minutes"—but they can keep flapping their wings long enough for the Harvard researchers to learn everything they need to know from each successive generation of bots. For commercial applications, however, the RoboBees would need to be more durable.
So what’s the end-game? Wood says he’s uncomfortable speculating on future uses for the RoboBee, which still needs many years of research before it can be commercialized. But he admits that one of the original motivators for the project was colony collapse disorder. Theoretically, the RoboBee could one day take over for a dwindling bee supply, pollinating crops. The other obvious uses for the RoboBee are in search and rescue (where it could find disaster survivors), where small robots are already common, and environmental marketing.
It also seems likely that someone will eventually use these things for spying. For now, Wood says that the vision sensors can’t deliver video feeds or images. For now.