Aircraft maintenance. Data is everything

Since 2013, the United States Air Force has lost 224 pilots and $ 12 billion to aircraft design and equipment problems. The National Military Aviation Safety Commission investigated the causes of these disruptions. Along with the psychology of the pilot and the lack of funding, the main problem is the lack of artificial intelligence (AI) technologies in servicing aircraft equipment.

According to the commission’s instructions, AI began to be used in the form of predictive maintenance of equipment (Predictive Maintenance). So the military was able to prevent accidents and failures, increase the safety of ships and minimize the cost of their maintenance – up to $ 5 billion annually.

In this article, we will discuss how military aviation has set the trend for predictive maintenance and how it helps to save on maintenance of the aircraft fleet.

From small details to millions in losses

The refueling tanker KC-135 of the US Navy broke down about 20 times, consuming $ 6.6 million from the budget. A detailed analysis of the reasons showed that each time the breakdown occurred due to one hydraulic pump, which regularly failed.

To end this Groundhog Day, the command changed the approach to working with equipment from reactive to predictive (Predictive Maintenance). Thanks to Predictive Maintenance, algorithms analyze data for every aircraft part and automatically warn about repair or replacement.

Lockheed Martin is also taking advantage of predictive approaches on Sikorsky helicopters. The company implements systems inside ships to search, structure and analyze information from different modules. Thus, she optimized the availability of the helicopter fleet and saved equipment maintenance costs.

CIS civil aviation also has examples of the use of predictive maintenance systems for equipment. S7 Airlines has implemented an EAM solution for the Airbus A319, which analyzes an array of data on component performance and meteorological factors. This information helps to reduce the number of flights delayed for technical reasons, predict breakdowns in the fleet and make flights safer.

Data is everything

In aviation, equipment data comes from various sources and in huge quantities – almost 100 million terabytes per year. Such arrays are called Big Data, and their main sources are:

– flight data recorders (FDR);

– engine condition sensors (EHM);

– sensors of the state of the onboard equipment (AHM).

Each module processes thousands of parameters down to the “health” of an individual component. To avoid losing this information, aviators need a powerful IT infrastructure.

This is how companies came to EAM systems that can quickly and easily collect data on aircraft equipment and predict when and what specific part will fail. Moreover, the system collects “data history” for the past periods, which means that every month the mathematical model becomes even more accurate.

Lufthansa even said that predictive maintenance solutions for equipment have reduced them 80% of unplanned downtime and helped to save about 1 million euros per year.

Predictive maintenance as an add-on

If you have read our material about equipment maintenance techniques, then you know that Predictive Maintenance is not a panacea, but the next step in the effective management of units. Eliminate the maintenance blind spots inherent in reactive and preventative approaches. For clarity, let’s compare each of them.

Reactive

Easy to implement and run.
There is no investment in IT (software, sensors and training of specialists).
Unpredictable downtime and overtime as a result.
Unexpected breakdowns and, as a result, more losses and a shorter asset life.

Preventive

MTBF – less unplanned downtime.
It is difficult to implement in a large aircraft fleet.
Requires additional (and optional) scheduled downtime and parts procurement – too expensive.
Forecast

Less downtime due to predicted breakdowns.
A complete picture of the “health” of the fleet in real time.
Fast repairs even for occasional accidents.
Especially important for critical systems such as jet engines and auxiliary drive modules.
Predictive maintenance is also mutually beneficial for other systems: knowing the condition of components in real time, it is easier and cheaper to react reactively to breakdowns; and having a real life cycle of an asset, it is much easier to implement preventive practices.

Total

Collecting, structuring and processing data play a key role in reducing the cost of servicing an air fleet. This has a positive effect on safety: you control not only airliners, but also ground support equipment.

Another factor is financial. Products like SmartEAM predict the failure of even critical components like a gas turbine engine. Thus, airlines save significant amounts on timely maintenance or replacement, rather than spending on major repairs after an accident. And given the development of the aviation and aerospace industry, predictive maintenance will help airlines not only save money, but also become part of the technology of the future.