A team of engineers at Harvard University has taken cues from Nature to create the first robotic fly. The mechanical bug has become a platform for a suite of new high-tech integrated systems.
A team of engineers designed a robot to do what a fly does naturally. The tiny machine is the size of a fat housefly. It’s agile and fast. Its miniature flapping wings allow it to hover in place and perform controlled flight maneuvers.
“It’s extremely important for us to think about this as a whole system and not just the sum of a bunch of individual parts," said Robert Wood.
Harvard engineering professor Robert Wood has been working on the robotic fly project for over a decade. A few years ago, his team at Harvard’s Wyss Institute for Biologically Inspired Engineering got the go-ahead to start piecing together the components.
“The added difficulty with a project like this is that literally none of those components are off the shelf and so we have to develop them all from scratch," he said.
They engineered a propulsion system with wings, tiny actuators to drive the wings, and a mechanism to maintain proper wing alignment.
“The seemingly simple system which just flaps the wings has a number of interdependencies on the individual components, each of which individually has to perform well, but then has to be matched well to everything it’s connected to," said Wood.
The flight apparatus was integrated into a set of power, computation, sensing and control systems. Wood says the success of the project proves that the flying robot with these miniaturized components can be built and manufactured.
While this prototype robotic flyer is tethered to a small, off-board power source, the goal is eventually to make it autonomous, so that it might someday perform surveillance and data- gathering work at rescue sites, in farmers’ fields or on the battlefield.
“Otherwise the fly is totally unconstrained. Basically it can take off, land and fly around," he said.
Wood says the design offers a new way to study flight mechanics and control at insect-scale. Yet, the power, sensing and computation technologies on board could have much broader applications.
“You can start thinking about using them to answer open scientific questions, you know, to study biology in ways that would be difficult with the animals, but instead using these robots," he said. "So there’s a host of technologies and open interesting scientific questions that are really what drives us on a day to day basis.”
Wood says that while he finds real flies to be annoying little bugs, curiosity and awe at their mechanics inspired his design. He and his colleagues describe their work in an article in this week’s edition of the journal Science.