Master Sgt. Aaron Cordroch, Master Sgt. David Nunez, Andrew Crafts, Dr. Joseph Toup, Tim Floyd, Brandon Fairfield, Tyler Payne, Clifford Walton, Lt. Col. Justin Hickey, and Stephanie Smith.

You’re driving down the road and hear a ding. You look down at your dashboard and see a little yellow light in the shape of an engine.

You’re uneasy, because you know it could be any of a number of things your car is trying to tell you – some of them easy to fix, some of them most certainly not. With most modern cars, you can take it to a specialist who plugs into the car’s computer and tells you exactly what the issue is.

Unfortunately, with aircraft, which are orders of magnitude more complicated than cars – and in many cases much older – it’s not quite as easy.

Until now, that is.

“Now we can see what the aircraft is seeing that triggers it to report that dashboard light. That includes all the communication and information that leads up to a failure,’” said U.S. Air Force Master Sgt. Aaron Cordroch. He is with the maintenance directorate at Air Force Special Operations Command at Hurlburt Field, Florida, and he is the de facto spokesperson for a team that is revolutionizing aircraft maintenance with data analytics.

It’s called AMATS. The Advanced Maintenance Analytics & Troubleshooting System. And in a short time it has gone from a proof of concept to a top budgeting priority at the Air Staff.  

It started about three years ago when Cordroch and his teammate, Master Sgt. David Nunez, were part of a cybersecurity initiative where they witnessed a capabilities demonstration on a twin-engined special operations aircraft called the C-146 Wolfhound. The demonstration highlighted how available data analysis hardware could "scrape the data" off an aircraft's computer communications to identify potential cyber threats.

This process also yielded gigabytes of extra data that was just "noise" to the demonstration team. To Cordroch and Nunez, it was a wealth of maintenance status information waiting to be exploited.

“If you push the throttle forward [we could see] the electrical signal that says move the throttle up. We can see everything,”  Cordroch said.

He had been put in contact with Andrew Crafts (whom he fondly refers to as the “forty-pound wrinkly old brain” that spurred the project). Crafts saw not just the potential of that data but the path forward to be able to use what was, up to that point, proprietary data to get actionable insights. At the time, Crafts was a noncommissioned officer in the communications squadron, and Cordroch pointed out that it was crucial to have organic Air Force talent taking on this project.

When they started to examine the data, they realized they were onto something unprecedented.

“We realized we were able to sample the aircraft’s communications at a much higher rate,” Cordroch said. What had once been 40 megabytes per flight was now six gigabytes per flying hour. That level of fidelity led to an early win demonstrated on a C-130J.

“There was a recurring error and the team couldn’t figure out what was wrong,” he said. “We plugged [the AMATS component called] Vampire in, and the dashboard lit up like a Christmas tree with errors that were all caused by one component.”

That component had never indicated problems, but it was having downstream effects on the rest of the aircraft and causing the fault. The team quickly realized this unprecedented level of insight would have a major impact on real world maintenance operations.

For example, there is a system on aircraft engines called FODS – the fire overheat detection system. The system works by detecting the inability of a signal to pass through, which it would in the case of components around the engine melting in extreme heat. The problem is, the overheat indicator does not tell maintainers which engine is in trouble.

“A FODS overheat usually takes up a six-hour day for a team trying to troubleshoot it,” Cordroch said, because the maintenance teams have to go looking across multiple engines.

Cordroch’s team plugged in the Vampire to the aircraft.

“Immediately, we could see which segment failed on which engine. The team was in and out in 45 minutes.”

This is the real impact of AMATS. Having clarity of what is causing issues means aircraft spend less time in maintenance status and more time available for their mission. It also means cost savings for the Department of Defense with fewer maintenance costs and work hours.

But it’s what’s possible with the large amounts of this type of data that the team is excited about. Cordroch recalled an example where Crafts was able to see the degradation of a part over time based only on the lag in communication speed, replacing the part long before it ever indicated a fault, or worse, caused a mechanical failure. That level of predictive analytics promises revolutionary changes to how the military approaches maintenance.

Good news doesn’t wait long. Once the team had the project in Spark Tank, it hit a series of quick investments, including from the system program office and Headquarters Air Force levels. AMATS caught the attention of organizations across the Air Force and Cordroch said the team, which in all has now involved eleven people, is already in discussions with the Navy.

Cordroch also said the Rapid Sustainment Office has gone “all in” on predictive maintenance for C-130s, with AMATS looking at upwards of $23 million of procurement funds to start to field the technology.

It’s an impressive distance to cover in such a short time.

About the team, Cordroch  said, “the passion for leaving maintenance better than we found it has been the driving force that has allowed this team to move mountains.”

The Advanced Maintenance Analytics & Troubleshooting System is the winner of VISION’s 2023 Blaze Award for an innovation making significant mission impact.