RoboBee Flies Untethered for First Time, Lightest Vehicle Ever to Do So

In a record-setting achievement, Harvard researchers have flown their RoboBee microrobot untethered for the first time, making the decades-old robot the lightest ever machine to accomplish the feat.

RoboBee Flies on Its Own For the First Time

Engineers at Harvard University have flown their decades-old Robobee robot untethered for the first time, according to a new paper published today in the journal Nature.

RELATED: LEARN TO MAKE ROBOTS WITHOUT KNOWING HOW TO CODE

Harvard postdoctoral fellow Noah Jafferis, also lead author of today's paper, and Harvard Ph.D. candidate and co-author Elizabeth Farrell Helbling had both been working on the RoboBee project for about six years when last August, they flipped on the high-intensity halogen lights that gave RoboBee's photovoltaic cells the energy it needed to achieve its historic flight.

"This is a result several decades in the making," said the RoboBee project's principal investigator Robert Wood, Charles River Professor of Engineering and Applied Sciences at Harvard's School of Engineering and Applied Sciences.

The challenge for the Harvard team is a familiar one for any engineering project, but especially for robotics: balancing the weight of the machine with the components that power it, whether it's a battery or a photovoltaic panel.

This challenge comes to dominate nearly any other once you get down to the scale of robots that weigh less than 500 milligrams, a field known as microrobotics. At this scale, the lightest batteries currently available would eat up more than half of this weight, making the actuators that convert the energy from the battery into the kinetic energy that moves the robot far too weak to be effective. 

Photovoltaic cells, meanwhile, don't weigh nearly as much, but smaller photovoltaic cells are still limited in the amount of energy they can produce, and so far this hasn't been able to generate the number of milliwatts needed to power the wing actuators for the RoboBee. This has meant that until now, RoboBee needed to be tethered to an external power source to fly.

Flight Requires a Lot of Power

"Powering flight is something of a Catch-22 as the tradeoff between mass and power becomes extremely problematic at small scales where flight is inherently inefficient," Wood said. "It doesn't help that even the smallest commercially available batteries weigh much more than the robot."

As a result, there didn't seem to be much more they could improve in terms of power generation. Using an array of the smallest solar cells commercially available, weighing about 10 milligrams each, the Harvard scientists had to find ways to get the most out of them. Even under the most intense sunlight possible, these cells only provided 0.76 milliwatts per milligram of power.

So, the team went back and started making several major changes to the design of the RoboBee, including an important addition of a second set of wings. The new RoboBee X-Wing, as the researchers took to calling the new design, proved to be much more powerful than its predecessors.

"The change from two to four wings," Jafferis said, "along with less visible changes to the actuator and transmission ratio, made the vehicle more efficient, gave it more lift, and allowed us to put everything we need on-board without using more power."

Weighing just 259 milligrams altogether--about a quarter of the weight of a standard paperclip--the researchers were able to use high-intensity halogen light bulbs to provide the RoboBee X-Wing with the power it needed to lift off and achieve sustained flight, according to their paper. At 120 milliwatts, the power needed by the RoboBee X-Wing isn't even enough to power a single lightbulb on a string of Christmas lights, but it is still three times as much power as their solar array could produce under the most intense sunlight. As such, the RoboBee, while untethered from external power, is stuck under artificial lights in a lab for the time being.

The microrobot also lacks any kind of onboard control mechanisms, so while it can sustain flight, it cannot control where it is going. This and other challenges remain for the RoboBee team, but after a decade of needing to be plugged in for it to operate at all, the team has still cleared a major hurdle.

"Over the life of this project we have sequentially developed solutions to challenging problems, like how to build complex devices at millimeter scales, how to create high-performance millimeter-scale artificial muscles, bioinspired designs, and novel sensors, and flight control strategies," Wood said. "Now that power solutions are emerging, the next step is onboard control."

In short, engineering at any scale is a never-ending process of iterative improvements tackling one challenge at a time, and RoboBee is no different.

"When you see engineering in movies, if something doesn't work, people hack at it once or twice and suddenly it works. Real science isn't like that," Helbling said. "We hacked at this problem in every which way to finally achieve what we did. In the end, it's pretty thrilling."