Eyes in the Skies

Wake Forest University scientists pioneer using drone aircraft for ecological monitoring. They're flying drones over the Amazon rainforest to study the tree canopy and the amount of carbon dioxide the forest releases.

WINSTON-SALEM - In a grassy field between classroom buildings at Wake Forest University in Winston-Salem, two men are huddled over a computer screen.

“I’ve got auto-stabilize,” says Max Messinger, a biology graduate student, who is looking over the computer screen. “Okay, now we’re in stabilize, and check that, the battery looks good. I’m arming up.”

Messinger aims the controller box he is holding at a device in the middle of the field. He flicks a switch. The propellers on the device start to spin. The high-pitched whine gets louder and the propellers turn faster.

“Armed, and I’m getting it trimmed out,” yells Messinger.

The device on which everyone is focused looks like a spider on steroids. Black legs all connect to a large body in the middle of the mechanical creature. The center section is packed with wires and other technical devices. Six propellers sit atop the legs. With a shudder, the spider-looking, high-tech, research tool begins to fly.

“And we’re off,” Messinger shouts while smiling. He built the drone for Wake Forest University’s Center for Energy, Environment and Sustainability.

“This is a technology that has matured over the past 10 or so years,” Messinger explains. “Our drone has six rotors, it has computerized motor controllers and that is really the brains of the craft. Also, there is the onboard computer that has all of the sensors built into it. Those sensors detect the angle of the aircraft and then send the information to the motor controllers. The controllers then relay signals to speed up or slow down the motors to keep it stable.”

Messinger continues his flying and his explanation.

It has a GPS and compass so it knows where it is in space and that allows it to fly preprogrammed flight paths autonomously, land and take off and that sort of thing,” Messinger concludes.

The opticoptor is made from carbon fiber and aluminum parts. It costs about $2,000 to build and the actual assembly only takes a few days. It takes several more weeks to fine-tune the craft.

Usually the drone is flying a preprogrammed route: after take-off fly to point “A”, then “B”, circle around point “C” then return to base. The controller watches a live video screen as the craft follows the pre-programmed route, ready to take control if something goes wrong.

But on this day, Messinger moves the toggle switches on the controller to guide the craft. How it actually follows the pilot’s directions is determined by the computer relaying the directions to the motor controllers, which then adjusts the speed of the rotors.  Direction, altitude, and even level flight are all determined by rotor speed. Two batteries power the drone, providing about 15 minutes of flight time. The top speed of the craft is about 20 miles-per-hour.

“I will tell it I want to go forward, and then the computer will convert that command to speed up these rotors and slow down those rotors, and that will tilt the aircraft forward and it will hold that, allowing it to move,” explains Messinger.

“This entire craft, this robot, is all designed to provide a platform for these,” says Miles Silman, a biology professor and Director of the Center for Energy, Environment and Sustainability, pointing to the bank of cameras hanging beneath the drone. “Down here is the camera mounted to the gimble stabilizer, so this has sensors to keep the camera stable in two dimensions.”

It seemed like a lot of work and technology to transport a camera, until I saw on a computer screen in Silman’s lab just what a drone is able to achieve.

The drone flying over the pond took the video of Duke Energy’s coal ash pond that spilled in the Dan River near Eden, NC. The craft flew a grid pattern over the pond, snapping roughly 250 images at regular intervals. The images were then compiled to create a 3-D model of the spill site.

“It’s fairly simple math once you have the images,” says Silman, as he moves the cursor around the image on the screen. “We can go and reconstruct the shoreline and then you use that as a surface, and since we know the depth of the pond, you can calculate the volume of stuff that is missing.”

By comparing the new image with existing aerial photos of the pond before the spill, researchers calculated about 35 million gallons of ash and water slurry from the pond. It provided independent verification of Duke Energy’s estimates. Researchers shared the data with Duke Energy, federal and state regulators and environmental groups.

It was data best captured by a drone.

“So when you think about the most widely used satellites, that gives a resolution that is about 90 feet on a side when you look at an image,” says Silman, who has used satellite images in previous research projects. “Those types of images work for many things. But if you are interested in finer detail, such as flowers or tree crowns, you need greater resolution. And that is what these robots allow as they operate in 3-D space. We can look at small scales.”

The team plans to use the drones to study the effects of climate change on the Amazon rainforest. Their target is the western Amazon region of Madre de Dios, in Peru. The region is about 3,000 miles wide and it is home to an estimated 390 billion trees. Because it is so vast, the rainforest is a major catalyst for Earth’s climate and weather patterns. The drones can fly close enough and steady enough to individual trees to allow for a study of leaf canopy temperature, which will help determine how much water vapor the trees will put into the atmosphere and how much carbon dioxide will be removed.

“We will be able to answer a lot of problems that were intractable and to think about new areas of research,” says Silman smiling, as he looks at the monitor and then glances over his shoulder at the spider-like drone sitting on the workbench. “We can do things with these drones that we were unable to do before.”

Related Resources: