Usually, when you dunk a tiny flying robot in the water you end up with a tiny sinking robot. Engineers at the Harvard John A. Paulson School of Engineering and Applied Science (SEAS) want to change that with the RoboBee, which has claimed the title of the first insect-insect sized robot that can swim as well as fly.

Building a machine that can operate both in the air and underwater is an exercise in contradictions. An airplane fuselage must be light, while a submarine's hull must be strong. An airplane needs to be narrow, while a submarine is best when shaped like a whale or fat cylinder. Above all, an airplane needs wings for lift, while a submarine needs to be smooth to reduce drag. Though various engineers have worked for decades to create flying submarines, the results have always been something that is neither a very good airplane, nor a very good submarine.


More than 1,500 New Atlas Plus subscribers directly support our journalism, and get access to our premium ad-free site and email newsletter. Join them for just US$19 a year.


However, nature solved the problem millions of years ago with seabirds that can dive underwater and swim for considerable distances before coming up with a nice fish for supper. It was to these examples that the SEAS team turned for inspiration – specifically, to the puffin, which changes from a flyer to a swimmer by adapting the flapping motion of its wings for water propulsion. By using theoretical, computational, and experimental studies, graduate student Kevin Chen says the team was able to determine that the only major difference in the two modes of propulsion is the speed that the wings move at. It was such a simple change that they realized that the Harvard RoboBee actually had the potential to both fly and swim.

Smaller than a paperclip, the RoboBee weighs only 80 milligrams and is made of flat layers of laser-cut carbon fiber hinged together with embedded plastic to form a frame and uses piezoelectric actuators to flap gossamer-like plastic wings at 120 beats per second. To save weight, power comes from an external source by way of a wire tether.

Ironically, it was this lack of weight that posed the first obstacle for the SEAS team. RoboBee is so light that it can't break through the surface tension of water, so any attempt to land and dive would result in the tiny robot flapping about on top of the water like a trapped aphid waiting for a hungry trout to pass by. To solve this, the team used the somewhat inelegant solution of setting the RoboBee at an angle above the water, switching off the power, and letting it crash with enough force to penetrate and sink.

At this point, the modifications made to RoboBee and its experimental water bowl came into play. Since the robot is electrically powered, the team had to prevent the water from shorting it out. For experimental purposes, the team used deionized water or purified water that's had all of its mineral ions removed. Contrary to popular belief, pure water is a very poor conductor of electricity. It's the mineral ions dissolved in it that actually carry current. By removing the ions, the team reduced the chances of a short. To better improve things, RoboBee was further insulated by coating its electrical connections with glue.

The next problem was one of density. Since water is almost 1000 times denser than air, the speed of the wings flapping had to be reduced – as per the puffin – to prevent them from being snapped off. In this case, from 120 beats per second down to nine. Steering was accomplished by altering the wings' stroke angle. The result is a flying robot that looks as at home underwater as a water bug.

The team says that the next step in RoboBee's development will be to find a way for it to safely generate enough lift to transition from water back to air to complete the circle.

"What is really exciting about this research is that our analysis of flapping-wing locomotion is not limited to insect-scaled vehicles," says Chen. "From millimeter-scaled insects to meter-scaled fishes and birds, flapping locomotion spans a range of sizes. This strategy has the potential to be adapted to larger aerial-aquatic robotic designs."

The SEAS team's results were presented in a paper at the International Conference on Intelligent Robots and Systems.

The video shows the swimming RoboBee in action.

Source: Harvard

View gallery - 3 images