Pilots fly UAVs in Iraq or Afghanistan, for example, through a web of fiber-optic cables and satellite dishes. Phone lines carry flight commands from US military ground control to satellite dishes in Europe, which then relay the commands to other small satellite dishes on the Predator, Martin explains.
The string of computer nodes means a two-second lag between the pilot’s command and the aircraft’s response. That’s why pilots based in, say, Kandahar fly the drones during takeoff, then transfer the controls to other Air Force pilots based in Nevada. When it’s time for the drones to land, the Nevada-based pilots then hand back the controls to their counterparts in Kandahar.
But given the complex and varied nature of the communications network that connects the pilots to their drones, links are often lost.
Martin recalls an instance in which his screen froze just has he was preparing to fire one of the missiles on his Predator. “Suddenly, just before I fired, the screen froze. Alarms began sounding and warnings flashed,” he recounts in his book.
“There had been a power outage in the satellite uplink in Europe. I had no further control of the plane.”
In such cases, UAVs are left to their own devices. Though they are generally programmed to fly a pre-determined – and thus trackable – path until pilots are back online, in the past the US military has been forced to shoot down renegade drones. In other instances, they simply crash.