Adam Wells Blog - Data Collection and Storage, Testing Today – Predicting…
Home » Blogs

How is it we put a man on the moon in 1969, but today we still can’t figure out when an aircraft electrical system is about to fail? Electrical failures are unique, as I’ve seen aircraft fail in flight on their first mission, even though they passed operational testing the day before. How can this be, since it just PASSED?

Because all fielded test sets report either pass or fail, the maintainer operating the test set only knows those two parameters, aircraft pass or aircraft fail. Well, to better understand when an electrical system will fail, you need to go deeper. With hardware it can be fairly straightforward. Brake pads wear at a certain rate: calculate “X” number of miles to achieve said wear, recommend changing pads at “X” miles. Did the brake pads fail? No, they just calculated a high percentage of brake pad material had worn off, and failure may be imminent, so they recommend changing the brake pads sooner rather than later.

Since electronics aren’t consumed, you can’t calculate their failure. So, when a pass/fail test set is currently used, it is only reporting system status at that exact moment in time and nothing else is known. However, if test results could be captured, compared and analyzed over time, predicting the needs of tomorrow would be possible. With the addition of these test-set test reports, values could be analyzed for trends and failures. AI software that’s continuously combing through stored reports, searching for new trends, drifting values and high-fail items, may be all that is needed.

Early in 2020, Marvin Test Solutions began working with an AI software company, and in the process we discovered the potential for creating a “predictive maintenance tool kit” that could be employed all over the world. Marvin Tests Solutions’ MTS-3060A SmartCan™, paired with powerful AI software, could become a predictive maintenance O-level test set, putting tomorrow’s problems in the maintainer’s hands today.

Maintainers who employed this solution would no longer be held hostage to conducting reactionary maintenance, due to snapshots-in-time of their aircraft’s electrical systems. Instead, they would be able to incorporate soon-to-fail items into their normal work flow, thus controlling the maintenance cycle and fixing potential failures before they happen.