James Junda stared down at the broken corpse of his drone. Just moments before, his five-pound helicopter had hovered beside an Osprey nest atop a 30-foot pole in northwestern Montana, its camera offering an up-close look at the adults and their two chicks. Apparently, Papa Osprey had tired of the spy. He batted the chopper, sending it spinning to the ground. The four blades snapped, an arm popped off, and the body broke into pieces. As Junda and an assistant, Marianna DiMauro, scurried along the rural road, gathering up bits of plastic and metal, the raptor circled above. “He was pretty proud of himself,” Junda recalls wryly. Indeed, victory was his: The drone—one of Junda’s two—was ruined.
But then, better the drone than Junda.
Light-aircraft crashes are the No. 1 killer of wildlife biologists. Between 1937 and 2000, 91 biologists and other scientists died in the field, according to a 2003 study in the Wildlife Society Bulletin, and 60 of them were killed in plane or helicopter crashes. What’s more, the study says, most of those 60 appeared to have been flying at the low altitudes necessary for observing and tracking wildlife. Recent years have seen more deaths. David Maehr, for instance, crashed and died in 2008 tracking radio-collared black bears in Florida, and Kristina Norstrom perished last year trailing caribou in Alberta.
Unmanned aerial vehicles (UAVs), known colloquially as drones, don’t offer just a safer way for scientists to observe their subjects; they’re often less costly, more efficient, and more precise than traditional approaches. As Junda can attest, biologists are still navigating the challenges and complexities of mixing wildlife and drones, working to improve maneuverability, autonomy, and endurance—and hitting regulatory hurdles. Learning curve aside, the technology holds incredible potential. “Drones are going to change the way that data is collected,” says Leanne Hanson, a U.S. Geological Survey biologist who has used them to count migrating Sandhill Cranes.
nlike the UAVs used by the military, most notoriously to kill enemy combatants in places like Pakistan and Afghanistan, the drones biologists employ carry cameras and sensors, not weapons. In the past few years they’ve started to provide an unprecedented look at hard-to-reach places, like orangutan nests high in the jungles of Sumatra and Borneo, and have shown potential for catching poachers and stopping illegal logging (see “Drone On” below). And because they carry digital cameras that produce geo-referenced photos, the data they gather can be fed into image-recognition algorithms to vastly improve the accuracy of population counts.
As more UAV manufacturers enter the field, the technology is advancing even as the price tag drops. 3D Robotics, for instance, has a suite of systems for $1,300 and less. The California-based company aims to create affordable drones for farmers, everyday consumers, and Hollywood, says the company’s editorial director, Roger Sollenberger. “But our drones are very adaptable.” In fact, ornithologists are using them to relocate California Condors that stray out of their original feeding grounds when searching for carrion. While the cheapest options usually involve buying a GoPro camera and a GPS, obtaining sophisticated data requires shelling out some extra cash. “If you want more than just a picture, then significant skill, expensive programs, and high-end computers are necessary,” says Franklin Percival, who leads the University of Florida’s Unmanned Aircraft Systems Research Program.
While biologists are incorporating UAVs into their fieldwork, drone researchers are designing newer models that are both more maneuverable and autonomous. Prototype insect-like drones are far less likely to break apart on impact (something Junda would appreciate). Vision-based navigation that allows the drone to automatically circumvent obstacles promises more precision than GPS-based systems. Going beyond eyes in the sky, researchers are attaching arms that grab objects in midair like an eagle, and kestrel-like legs that allow drones to perch. They’re building planes that glide like albatrosses by exploiting thermal updrafts, and developing solar-powered drones that never have to land and refuel. One group is even building drones that mimic flocking birds, moving in unison without crashing into one another.
Despite the progress, there’s a major hitch. Red tape is keeping the fledgling field of drone ecology from getting off the ground. “Right now the law says my little 5-pound drone is equivalent to a 50-pound, gas-operated fixed-wing aircraft, and it’s just not,” says Junda. “We need better regulations.”
Federal Aviation Administration (FAA) rules currently make it difficult and time-consuming to obtain permission to fly drones outside. Would-be operators fill out a 20-page application; even if a group receives a certificate, they can’t fly in cities or other crowded places, and they are limited to a 20-square-mile area (to expand, they must get separate permission). What’s more, each flight requires a licensed pilot and an observer. In Canada and the United States, only public universities and government entities have received the green light to fly small drones (less than 50 pounds), with one exception: The FAA granted ConocoPhillips permission to monitor marine mammals and icebergs around its oil rigs in Arctic waters last summer.
Those restrictions could change soon. In 2012 Congress directed the FAA to craft guidelines to safely integrate small unmanned aircraft (such as those mentioned in this article) into the national airspace by September 2015. The regulations for small drones, which most wildlife biologists favor, should be out this year, an FAA spokesperson explained to Audubon in an email. These low-flying aircraft don’t pose a threat to commercial or military airplanes, though they could be a danger to smaller manned planes and helicopters that fly at lower altitudes. The government expects that the vast majority—80 percent—will be used in agriculture and 10 percent for domestic surveillance; the rest will take to the skies for everything from monitoring wildlife to tracking sporting events.
Adding to the chaos in this up-and-coming branch of aeronautics, a judge ruled in March that the FAA doesn’t have the rules in place to prevent operators from flying small drones for commercial purposes. “It could temporarily open the floodgates,” says David Bird, editor of the Journal of Unmanned Vehicle Systems and Junda’s master’s thesis supervisor. If someone were seriously injured or killed by an unpermitted operator, he says, “that would set the whole drone industry back a decade.”
Other countries have looser permit requirements, but operators still face hurdles. In Quebec, for example, there’s only one guy who reviews applications, says Bird. “He’s also responsible for balloon festivals and airplane shows.”
The resulting backlog is the reason Junda—who lives in Montreal—ended up in Missoula, piggybacking on the project run by Erick Greene, a drone-flight-certified biologist at the University of Montana.
n late July, Junda’s luck ran out with his second drone. He’d been flying over several Osprey nests daily for three months, taking pictures as the eggs gave way to chicks that grew astonishingly fast on a steady diet of fish. Junda is studying the safety of UAVs as raptor-monitoring tools, and whether Ospreys defend their nests more aggressively at certain times. He’s convinced the technology will provide more accurate population data, make the research easier, and drastically reduce monitoring costs. It’s far less cumbersome than using a bucket truck to see inside the nest, and, at $7,000 per drone, way cheaper than renting a full-sized helicopter—such a pricey endeavor it can limit fieldwork to a single day per season.
Still, Junda has no illusions that there are not still kinks to work out. One morning the second Draganflyer X4’s landing gear caught on a stick and flipped inside a nest. “I had this sinking feeling,” he says. “I thought I might have hurt the chicks or scared the parents and they might abandon the nest.” Instead, as captured by the still-recording video camera, the male whacked the helicopter with a stick a few times before placing it on top of the X4, effectively incorporating the drone into the nest. A guy on the university logging team donned spikes and recovered the seemingly undamaged chopper. But a month later it suddenly broke apart while descending. Junda once again scooped up the fragments. He moved pieces around, ran diagnostic tests, and spent hours on the phone with Draganflyer technicians, to no avail. His fieldwork was over.
The following day, Junda and an assistant conducted some final research the old-fashioned way. On a ranch overlooking the Bitterroot River, they set up camp chairs and tallied how often Osprey nests were exposed to human activity and predators. Though frustrated by the previous day’s events, Junda was now more matter-of-fact. “Crashing is just part of the deal,” he says. “And it seems like a very fitting way to end it, sending back my two broken UAVs with a bag full of bolts and shattered pieces.”
By the next nesting season, Junda’s drones will be ready for flight. Maybe, just maybe, that lone agent in Quebec who approves flight permits will have gotten around to processing Junda’s application by then.
This story ran as "Eyes in the Sky" in the July-August 2014 issue.
UAVs are aiding wildlife research across the globe. Here’s a small sampling of projects.
Project: Surveying orangutan nests in Indonesia
Drone: Self-built drone made from a model airplane, autopilot software, and a camera; less than $2,000
Description: To find orangutan nests, biologists Serge Wich and Lian Pin Koh’s team fought through the dense Indonesian rainforest. Three years and $250,000 later, they had found a number of nests. But the conservationists needed a less costly approach, especially since illegal palm oil plantations and logging are rapidly destroying the apes’ habitat, says Wich. So they built their own fixed-wing plane. It brought back high-resolution video of the nests. The two biologists have since founded ConservationDrones.org, to help other scientists build UAVs.
Project: Monitoring Sandhill Cranes in Colorado
Drone: Raven RQ11A fixed-wing plane; $0 (hand-me-down from the U.S. military)
Description: Leanne Hanson, a biologist with the U.S. Geological Survey, compared drone and ground surveys of migrating Sandhill Cranes. The Raven surveys were more accurate, with the added benefit that the plane could fly over the resting flock in pre-dawn hours, when the birds are less likely to flush. “The images were amazingly clear,” Hanson says. “We saw roosting cranes that our ground counters missed.”
Project: Counting waterbirds in Florida
Drone: NOVA 2.1 fixed-wing plane; $200,000 for three planes and a ground station
Description: Getting an accurate count of nest- ing waterbirds is a daunting task: Some colonies have 20,000-plus nests in a two-mile stretch. UAVs built by Franklin Percival’s team at the University of Florida’s Unmanned Aircraft Systems Research Program fly a fixed route, snapping thousands of photos each with a unique GPS stamp. Back in the laboratory a powerful computer stitches the photos together, allowing scientists to pinpoint every bird. “How many of a particular species there are is the most difficult question biologists have to answer,” Percival says.
Project: Surveying seals in Arctic waters
Drone: ScanEagle, a fixed-wing UAV; $500,000 for the plane, navigation system, and operating expenses
Description: The ideal time to count seals in the Bering Sea is when they haul out of the frigid water onto sea ice. The problem? “It’s very hard to distinguish between what’s a seal and what’s a crack in the ice” with the human eye alone, says Elizabeth Weatherhead, senior scientist at the University of Colorado. Computer analysis of the 27,000 photos taken by ScanEagle, built by a Boeing subsidiary, revealed seals Weatherhead says she never would have spotted.
Project: Counting Chinook salmon nests in Oregon, Washington, and Utah
Drone: MikroKopter, hexacopter; $8,000
Description: For years Phil Groves helicoptered into Hells Canyon, North America’s deepest chasm, to hover 300 feet above the water and count the nests of endangered Chinook salmon. After some colleagues died nearby in a 2010 helicopter crash, he convinced his employer, Idaho Power, to buy a drone to conduct the surveys. “It was an absolute, clear success,” he says. “Then the FAA called us and said, ‘You can’t do that anymore,’ ” because the company isn’t a public entity. New regulations expected later this year could allow him to resume the project.