Do Drones Get Vertigo, Too?

Up there or down here, it can be a struggle to maintain “situational awareness.”

NASA's IKHANA research craft is a modified Predator B. NASA

In a roadmap for the next 25 years of unmanned aerial vehicles (UAVs), published in 2002, the Department of Defense made it pretty clear who it regards as the weak link in military aviation: pilots.

“A UAV will never be lost because a pilot experiences vertigo,” the roadmap’s authors wrote. And the list of human frailties did not stop there. “No UAV is likely to be lost due to aircrew urgency to return to base and family. Robotic aircraft are not programmed to take chances.”

A year after the report, two Air Force Academy cadets, Elizabeth DeLauer and Cory Fale, decided to test the assumption that, unlike onboard pilots, human UAV operators are immune from the effects of vertigo. They studied 10 military pilots as they flew a UAV in a simulator at constant rates of climb and descent, and while banking at 30 degrees. Then they altered the motion of the platform from which the flights were conducted, to simulate the effects of flying a UAV from a moving aircraft such as a C-130 Hercules. “Even though the…operator is not located in the vehicle, he or she can still experience spatial disorientation,” they concluded. The most severe effects on control, particularly on pitch (while climbing or descending), occurred when the pilot’s platform was flown in a different plane than the UAV's. The researchers' conclusion: Best to leave the pitch on autopilot.

Few unmanned vehicles are piloted from a moving vehicle. Typically they’re flown from a trailer or even an office, thousands of miles from the aircraft. No unmanned aerial system, or UAS (the term preferred by the Department of Defense and the Federal Aviation Administration, since it includes the human operator and the data link), has been lost solely to vertigo. But several have crashed or strayed due to human error. In 2006, a Predator B UAS flown remotely by U.S. Customs and Border Protection lost power before dawn and crashed near the Mexican border. Its operator, toward the end of a dull overnight shift, failed to follow a checklist when switching control from one display panel to another, inadvertently shutting off the vehicle’s fuel.

Last April, the National Transportation Safety Board convened panelists to discuss whether unmanned vehicles could happily share the nation's airspace with piloted airplanes. The NTSB heard examples of UAS operators losing “situational awareness” due to factors ranging from fatigue to insufficient training.

Forum panelists said it's unlikely a remote UAS operator would experience vertigo that would result in his UAS colliding with an airliner. But collisions with powerlines, cell phone towers, or a skyscraper were a worry. Most panelists agreed that a UAS pilot should meet the same medical requirements and undertake the same vertigo training as pilots of comparable conventional aircraft, and that they should limit their duty time and be required to take rest periods, just like airline pilots. On the civilian side (for example, UAVs used in police work), safeguards in the Federal Aviation Regulations limit the duty time and ensure rest periods.

Kevin Williams of the FAA Civil Aerospace Medical Institute has worked for medical standards for UAS pilots. During the NTSB panel discussion, he pointed to several disorienting factors that affect UAS operations, including latency, or the time delay in the relaying of information between vehicle and pilot. The camera system on the Pioneer UAS retracts just before landing to protect it from a rough touchdown, at which point the remote pilot loses all video. A spotter at the remote runway has to radio the operator to shut down the engine after landing; in one instance, an ambiguous call led to a premature shutdown and a crash.

An authority on training conventional pilots to handle vertigo, Williams said that UAS pilots lack the vital sensory clues that warn of an impending risk. These include changes to the ambient noise, the sensation of pitch and yaw, vibration, and control forces becoming heavy or mushy. Even the sense of smell is important as an indicator of trouble in a cockpit. All such “diagnostic” cues are absent in the UAS. “You can simulate some of this information, but you can’t replace it,” said Williams, noting that 26.5 percent of accidents in the Predator were at least partly due to pilot misperceptions.

Mark Pestana serves as the remote pilot of NASA’s IKHANA vehicle, a Predator B the agency uses in UAS research. Pestana said that when it comes to sensory cues, UAS pilots are severely hobbled. “Imagine stepping into a cockpit and losing four of your five senses; you’d only have vision.” Control inputs for IKHANA are made by a series of trackball and joystick movements. “There was probably not a pilot in the room when this software was written,” said Pestana.

Nancy Cooke of Arizona State University studied the work style of UAS teams and told the NTSB that the lack of sensory input creates spatial disorientation, which in turn causes most mishaps, particularly while landing. “The visual experience of a UAS pilot is looking at the world through a soda straw,” said Cooke. At best, a UAS pilot sees a 30-degree field of vision.

Glen MacPherson, the U.S. Air Force chief of human factor studies, said that UAS accidents stem from the arcane and exhausting pilot-machine interface. “Fatigue has been expensive,” he said. “Predator crews who are ‘tele-operating’ in Iraq are at least as fatigued as those actually deployed in country.”

Should a UAS pilot become disoriented, Vertigo Inc. may have a solution. Vertigo has built a recovery system for UAVs weighing 20 to 1,600 pounds: a round parachute or steerable parafoil that deploys whenever the vehicle assumes an unusual attitude or encounters high G forces.

Get the latest stories in your inbox every weekday.