Most Active Stories
Crashed Air Force drone was flying with gear that couldn't handle cold
Besides some clouds on Nov. 12, 2013, Air National Guard crew members recalled to accident investigators that it was a nice day for flying.
But it was a cold, late fall day, just below freezing on the ground at Fort Drum, where the Air National Guard launches its remotely piloted aircraft for training missions over upstate New York.
A review of the Air Force's publicly released report into the accident has concluded the unmanned military drone that crashed into Lake Ontario that day was flying with equipment that didn't always function properly in cold weather.
The MQ-9 Reaper had been in the air on its training mission for nearly two hours, under the control of 174th Attack Wing pilots at Hancock Airfield in Syracuse, when it began to experience problems.
"It was kind of a perfect storm how this thing happened," said Air Force Col. Dana Hessheimer. He was president of the Accident Investigation Board that looked into the crash.
The board's nearly 700-page report was finished earlier this month and obtained by WRVO through a Freedom of Information request.
In it, the Air Force cites multiple navigation system failures for the crash, not pilot error.
Two of the three navigation systems on board, known as EGIs, failed and the pilots on the ground lost contact with the aircraft.
That sent it into an emergency, pre-programmed flight, flying circles far over Lake Ontario.
A launch and recovery crew based at Fort Drum then tried to regain control of the aircraft. Hessheimer says they did, but just as the pilot entered a right turn, he said the plane’s autopilot “zeroed out,” or stopped functioning momentarily.
"If he had waited another 20 seconds, it would of probably came back and then it would of flown back just fine," he said.
Instead, the Reaper inverted upside down and ended up in a flat spin.
"Kind of like, if you remember 'Top Gun,' when Goose died? That’s kind of like what the airplane was doing," Hessheimer said.
"I can't recover it," the pilot is quoted as saying in the transcripts.
The aircraft plunged into Lake Ontario, falling at a rate of 83 feet per second. A few pieces of wreckage washed up on shore two days later, but most of the plane was never found.
MQ-9s carry three navigation systems, in case one fails. The crash wasn’t the first time that particular aircraft had navigation problems for the 174th. Twice in the weeks before the accident, maintenance crews removed and replaced EGIs.
"Yeah, maybe a little bit of a red flag for it," Hessheimer said.
"Pretty darn cold"
Hessheimer wrote in the summary of his report that he also considered if air temperature played a role in the crash, but he wrote there was not enough evidence to say it was a "substantially contributing factor."
However, during a conference call, reporters pressed Col. Hessheimer on the issue. After a long pause, he said, "No, I can’t rule it out."
He wouldn’t elaborate, other than to say it was a software problem.
"I can't get into that," he said.
At 18,000 feet, where the drone was flying that November afternoon, the air temperature hovered around minus 34 degrees Fahrenheit.
"Pretty cold. I would think that’s pretty darn cold," said Agamemnon Crasssidis, an engineering professor at the Rochester Institute of Technology and the academic leader for upstate New York’s private drone research consortium, NUAIR.
Aircraft must be able to adjust to air temperatures, according to Crassidis, or they won’t perform properly.
"Everything’s temperature dependent," he said, "because that changes the density of the air around it."
Cold weather could have impacted the navigation software in two ways, according to Crassidis. One, it could have slowed the system’s ability to run the processors fast enough to operate the software. Or, the EGI couldn’t properly compute air temperature information in order to calibrate the aircraft’s movements.
The MQ-9 is manufactured by General Atomics. The company tested the two navigation systems previously pulled from the drone that crashed as part of the contractor report section of the accident investigation.
General Atomics exposed the EGIs to cold weather flight conditions. One of them failed “multiple times,” the report said. (The exact temperatures the EGI failed at is redacted from the report.)
A General Atomics spokeswoman said the company doesn’t comment on accidents.
The Air Force also acknowledges in a deficiency report -- which is entered into a military database of equipment issues -- dated Dec. 2, 2013, the version of the Reaper’s navigation system the 174th's MQ-9 was flying with can fail at cold temperatures.
[...] EGIs experience a loss of communication failure that is impacting flight safety, mission readiness and causing mission failures. This failure typically occurred in flight at cold temperature when the EGI stops updating [...]
A software upgrade
Honeywell Aeronautics, the maker of the navigation systems, declined to answer specific questions. However, spokesman Scott Sayres says they provided a software upgrade for the navigation systems in June 2012, more than a year before the 174th’s Reaper crashed.
"We provide the upgrade to General Atomics," he said. "If it was implemented or what they did with it, you’d need to talk to [General Atomics] about that."
Col. Hessheimer, with the Air Force, says the problem was identified within days of the accident and the Air Force sent out a world-wide recall for all MQ-9s.
"It’s kind of like G.M. lately," he said. "They do a recall. They did the same thing in the Air Force. It’s called an alert bulletin."
It ordered different versions of the software be installed.
Both the Air Force and General Atomics say running a different version of the software has prevented further cold-weather flight problems since the 174th's MQ-9 fell out of the sky that afternoon last November.
"I'm a huge proponent of these things and I know the benefit of them," Hessheimer said. "I know it's a new technology and it scares the public because people don't understand it."