Hank sat impassively on a Virginia Tech athletic field, ready to take it on the chin for the future of drone commerce. About 30 yards away, an eight-rotor unmanned copter hovered, buzzing like a swarm of bees. The 21-pound drone tilted forward, accelerated sharply and slammed into Hank’s head, smacking the crash-test dummy’s neck backward and embedding shards of shattered propeller in his plastic face.
There is little disagreement that the small- and medium-sized drones flooding the U.S. market can seriously injure or even kill someone. Understanding and minimizing the risk will be key to convincing regulators to expand their permitted uses.
“What we need to understand, really, is at what level does injury become death?” said Mark Blanks, director of the government-approved drone test center based at Virginia Tech’s Blacksburg, Virginia campus. “When does the threshold cross an unacceptable level?”
“So many people are watching these studies,” said Earl Lawrence, director of the Federal Aviation Administration’s Office of Unmanned Aircraft Systems Integration. “FAA needs it to support our rulemaking activities, but so does every other civil aviation authority and interest groups throughout the world. People want some answers.”
Only by testing can officials determine whether there’s a threshold of weight and design properties under which flights would be safe. And it paves the way for the use of commercial drones weighing up to 55 pounds, which are needed for deliveries and other business uses, but could pose a hazard if they fly off course or their batteries run out mid-flight.
Drone flights over crowds remain controversial and the subject of intense debate. The Academy of Model Aeronautics, an umbrella for clubs around the country where people fly remote-control aircraft, has no plans to allow such flights, President Richard Hanson said. “We think it’s too risky,” Hanson said.
While the research is still under FAA review, there are early indications of at least one piece of good news for the industry: When small consumer drones made of plastic strike an object like a human head, they tend to break apart, lessening the impact, according to David Arterburn, a researcher at the University of Alabama in Huntsville.
Arterburn heads the FAA’s research effort to determine how badly a drone would hurt a person, and whether it’s possible to create a class of vehicle that’s so light and soft they aren’t a hazard. The group conducted its crash tests on dummies last summer at Wichita State University in Kansas.
SZ DJI Technology Co., the world’s largest drone manufacturer, last week released a study arguing that craft weighing up to 4.9 pounds (2.2 kilograms)—which includes its best-selling Phantom models—pose minimal risk to people.
Time Warner Inc.’s CNN is conducting its own research on how to safely operate drones over crowds for news video photography, said Greg Agvent, senior director of CNN Aerial Imagery and Reporting.
The network is so far the only company to receive an FAA waiver to fly over the public with a small drone copter that is tethered to the ground to restrict its movements. It’s working to get broader permission, Agvent said.
One of the key ways to reduce risk is to enclose drone propellers so they can’t cut people, he said. A test at Aalborg University in Denmark showed a drone propeller could impale a side of pork—a stand-in for human flesh—like a dagger.
CNN is looking at a combination of other protections, such as parachutes, and ensuring its drones are more reliable, Agvent said.
“I want to use these unmanned systems to cover the news, not be the news,” Agvent said.
The tests at Virginia Tech are a joint effort between the Mid-Atlantic Aviation Partnership, an FAA-approved drone test site, and the school’s Institute for Critical Technology and Applied Science, which studies head injuries in car crashes, sports and the military. While it was not requested by FAA, researchers said they planned to make results available to the agency.
“Hank,” a dummy dressed in a VT engineering t-shirt, was equipped with 15 sensors in his head and neck. He was placed in a chair on the 50-yard-line of an indoor football practice facility.
Pilot and test engineer Andrew Kriz then fired up a DJI Phantom 3 quad-copter, which is one of the world’s most common models and weighs about three pounds. Kriz made a few practice runs, then flew the white plastic craft into Hank’s face.
The impact damaged the Phantom’s props, cracked the frame and tore off its camera. Graphs displayed on a computer showed spikes in energy as the craft hit Hank, but more analysis and additional tests will be needed to determine how badly it would hurt a human.
Next, Kriz aimed a DJI Inspire, a higher-end device weighing about eight pounds, at the dummy. On the first attempt, the drone hit the dummy’s chest and bounced off. Still flyable, the drone struck the dummy’s face and neck on a second try.
Finally, an eight-rotor DJI S1000, a larger craft weighing more than 20 pounds that professional photographers might want to use over a crowd, was flown at the dummy. Even though the dense center of the drone missed, arms holding its rotor blades struck the dummy’s face, rocking the neck back and forth. Two slivers of black propeller blades penetrated the dummy’s cheek.
As drones buzzed nearby, Blanks said the purpose of the experiments is to open a new frontier in drone operations. But it also highlights a truth about existing models:
“If you’re operating a large, multi-rotor aircraft in really close proximity to people and treating it as it’s a toy, then I think it’s asking for some pretty severe outcomes.”