Originally published on Connected Aviation Today this article explores how maintenance issues before or during a flight can severely impact passenger trust and brand loyalty. Continue reading to discover how airlines can proactively handle possible issues, improve safety, and enhance overall passenger experience while also getting major operational and financial benefits with predictive maintenance tools.
Few scenarios can impact a passenger’s trust in an airline and brand loyalty like a maintenance issue before or during a flight. Maintenance issues that delay flights can be massive inconveniences. Maintenance issues that require passengers to deboard a flight before it takes off can raise concerns about the safety of an aircraft. And maintenance issues that require emergency landings or planes to be rerouted can shake a passenger’s loyalty for a particular airline.
I know from experience because I’ve been on an aircraft that needed an emergency repair before takeoff.
As a college graduation gift, my mother planned a round-trip cruise to Cuba in 2019. Like most passengers, I expected my flight experience to go seamlessly. However, upon standing at the passenger gate to board my flight to Miami, I saw everyone pointing at the aircraft in shock. To my dismay, the left aircraft engine started smoking, and a dark cloud quickly blanketed the airport window. Airport staff then informed us that we would be delayed three hours for emergency maintenance.
Often, aircraft-on-ground (AOG) events like this are unavoidable. Like all machinery, aircraft require maintenance and repairs. Parts will eventually need to be replaced due to normal wear and tear. Unfortunately, emergency maintenance can disrupt flight plans and have a snowball effect of negative impacts on airport operations and passenger experiences – such as passengers losing confidence in that airline, all connecting flights getting delayed, and a halt in airport operations.
Luckily, with new data analytics and advanced artificial intelligence (AI) solutions, airlines can benefit from innovative prognostic health monitoring (PHM) solutions that turn data into predictive intelligence – enabling airlines to make proactive decisions that may reduce emergency repairs and maintenance delays.
A recent webinar by Collins Aerospace, “Enhancing Aircraft Maintenance: Past Analysis, Present Acceleration, Future Predictions,” explored ways that modern airlines can use historical and industry data to improve maintenance in the present and predict problems in the future through predictive maintenance solutions.
During the webinar, Brian McHugh, Sr. Manager of Predictive Maintenance and Data Analytics, Connected Aviation Solutions, shared a personal experience of a maintenance event on an aircraft and how a new custom self-service PHM solution could have been a valuable solution to prevent the event from ever happening and expedite the aircraft maintenance process.
Fire alert causes emergency landing
According to McHugh, “When I [was working at an airline], I had an A 320 take off. Shortly after takeoff, a bleed leak in the pylon resulted in a fire alert for the crew. [This caused an] emergency landing. Fire trucks rolled. Passengers were left very disconcerted.”
After looking at the historical flight data and comparing that data to logbook actions taken by the maintenance crew, McHugh realized that the temperatures continued to rise flight after flight within the aircraft machinery. “[I] realized a pressure release valve (PRV) was replaced about a week before that event. So, the mechanic didn’t [secure or torque] it correctly, and the vibration from the engine was causing that clamp to come off over time until it finally came loose and created that alert,” McHugh explained.
If the steadily increasing temperature in the pylon had been identified before it became an issue and triggered the fire alert, an emergency landing could have been avoided. Instead, the flight was diverted and forced to land at an airport where fire crews were waiting – an experience that invariably impacted how the passengers on board felt about the airline and the safety of its aircraft.
Like all machinery, aircraft require maintenance and repairs. Parts will eventually need to be replaced due to normal wear and tear. Unfortunately, emergency maintenance can disrupt flight plans and have a snowball effect of negative impacts on airport operations and passenger experiences…
This is where PHM solutions can play a role. If an acceptable pylon temperature was set in the solution, and alerts set to trigger should that temperature be exceeded, flight maintenance crews would have been alerted to the steadily increasing temperature in advance, and the aircraft could have been repaired before it flew again.
Unfortunately, data analytics and the creation of algorithms and alerts are often the job of data scientists and AI professionals – not aircraft maintenance professionals and engineers. This means that the people who know the most about repairing aircraft and the maintenance issues that impact aircraft are not always the ones building the algorithms and setting the alerts.
But, thanks to a new generation of PHM solutions, that’s changing.
Taking the power back
In the past, the creation of complex algorithms, AI tools, and other advanced data analytics solutions required the help of a data scientist who would program complex code within Python—a programming language used for automating tasks, such as flight planning or performance monitoring on aircraft—to allow aircraft engineers to get maintenance alerts to identify further PRV valve issues in the future and wait for that aircraft to go through the long maintenance and repair process before it could get back in the air.
New tools are giving the engineers – who know the most about the aircraft – the tools they need to become more predictive and proactive in their repair and maintenance operations without having any programming knowledge.
Through the low-code/no-code PHM tool within Collins Aerospace’s Ascentia analytics services, engineers can utilize the analytic developer studio to create custom threshold-based alerts to identify and predict maintenance issues before they occur.
According to McHugh, “Once you’ve identified [an] event, such as the lead leak in the pylon, for example, you can go into the custom analytic tool, spend about a minute [developing] that logic, and then you can quickly deploy it to the platform. It’s similar to going into Excel, [creating] a quick formula, and deploying it.”
Unfortunately, data analytics and the creation of algorithms and alerts are often the job of data scientists and AI professionals …This means that the people who know the most about repairing aircraft and the maintenance issues…are not always the ones building the algorithms and setting the alerts.
Analytic developer studio enables the capability to create private alerts or have those alerts viewable to the rest of the airline to manage maintenance workflows, so nothing is missed. This equips airlines with the ability to be more proactive and predictive in their aircraft maintenance and eliminates the need to wait for emergencies to happen.
“I had to wait for that PRV failure to happen, and that’s less than ideal. Airlines want to be more proactive in their maintenance… What Analytic Developer Studio [does] is find every pattern in your data that could possibly be turned into a predictive alert,” McHugh said.
Utilizing predictive data not only allows airlines to prevent maintenance events but also brings huge cost savings to airlines and minimizes AOG events by allowing airlines to plan aircraft maintenance before aircraft are scheduled to fly.
“[You’re not only preventing the maintenance event], but you’re also preventing the delay and the cancelation when the event doesn’t occur. You’re increasing safety, reducing the workload on your pilots, and turning an unscheduled event into a scheduled event,” said McHugh. “You don’t have to pay overtime for the maintenance crew [or] overnight shipping [for] parts… Everything goes so much smoother when it’s planned.”
“You’re also looking at reducing your repair costs, [having a] positive PR brand reputation [by reducing delays], and there’s also a lot of opportunity around fuel savings,” McHugh explained.