Airlines use predictive maintenance to slash costs, convert unscheduled maintenance to scheduled events and reduce asset downtime. They have been doing it on engines for some time, and now airlines can perform predictive maintenance much more effectively because new aircraft yield more data, and the tools for exploiting it have improved dramatically. Yet ways of thinking and business processes also must change, which often is more difficult than tapping sensors. And who will conduct the analyses upon which crucial maintenance decisions are made: OEMs, maintenance providers, airlines or consultants?
Most predictive maintenance today starts with OEM tools. Alan Epstein, vice president of environment and technology at Pratt & Whitney, describes predictive maintenance as using sensors and computers to do what human inspectors once did by disassembly and visual inspection. “Maintenance was always on condition; we just use computers to understand condition now,” he says.
Engine monitoring began when statistical correlations were observed and then traced to their physical causes.
Most Pratt customers use the company's Advanced Diagnostics and Engine Management (ADEM) to help plan engine maintenance. Epstein predicts Pratt and other engine OEMs will continue to improve prognostics, partly because their long-term support contracts offer huge incentives to keep costs down and availability up.
Auxiliary power units also are very mature in predictive techniques, notes Kristen Law, director of mechanical maintenance strategy and condition-based maintenance at. The company maintains a website showing trend data for its air-transport APUs and estimates how many hours are left before major repairs. This Predictive Trend Monitoring and Diagnostics (PTMD) tool offers troubleshooting tips, with estimates of the probability of each tip's success.
Honeywell also makes central maintenance computers to collect data from many subsystems. Its CMC on thecollects data from more than 80 systems and sends it to airlines or vendors. This CMC can intelligently filter multiple faults that propagate from a single failure and focus attention on the root of the problem.
But predictive possibilities differ from component to component. For example, for engines and APUs, systems can suggest maintenance actions, but only trend data is offered and flagged green, yellow or red for many other components. Someone at the airline or elsewhere must then decide what to do.
Law says one improvement would be giving ground staff the ability to request more data from onboard systems automatically, without talking to pilots. “But there are certification issues to sending messages in flight,” she notes.
Airframe OEMs have joined engine makers in monitoring aircraft performance.'s AIRcraft Maintenance Analysis (Airman), used by 106 customers, constantly monitors health and transmits faults or warning messages to ground control. The tool offers rapid access to maintenance documents and troubleshooting steps prioritized by likelihood of success.
Airbus's new Real-Time Health Monitoring (AiRTHM) goes further as systems on new aircraft yield more parameters, and the Aircraft Communication Addressing and Reporting System (Acars) enables Airbus to collect and analyze data from remote locations in real time.
It is developing AiRTHM in its Airtac (Airbus Technical Aircraft-on-ground Center) maintenance control center to give real-time troubleshooting assistance, guide spare provisioning and monitor system health to anticipate failures. It is available to ease the's entry into service, and it will be extended to A380 Flight-Hour-Services customers as well as supporting the .
's Airplane Health Management (AHM) is used on 2,000 aircraft for 53 customers. Dave Kinney, associate technical fellow in commercial aviation, describes AHM as “part of one pathway toward predictive maintenance. We see three primary elements: experience such as knowledge of aircraft design; tools like AHM and other analytical tools; and data, including operational and maintenance data.”
Looking at multiple data sources yields richer possibilities for prediction, and flight-data recorders capture and download huge data files in flight or on landing. Kinney says both physics-based and statistical methods are necessary for the best predictions.
One emerging tool, associative memory, maps connections between text and data from many different systems, much like looking for connections in the human brain. “Advanced analytics applied to more disparate data sources—that's how we are going to crack this,” Kinney says.
But a culture change is also necessary. “Mechanics are not used to removing a component because experts predict it will fail next month,” he notes. “This culture change is still in its infancy, even with high-value assets like engines.”
Predictive capabilities have become table stakes, even for regional aircraft makers.offers Ahead, an integrated tool that consolidates aircraft data from onboard systems and Web-based databases, to monitor and recommend maintenance for . The latest version, Ahead-PRO, will be available on all future commercial aircraft, according to Luiz Hamilton Lima, vice president of services and support.
About half of E-Jets and 40% of their operators use Ahead. Both Embraer and customers receive data, some via Acars in flight, with larger data sets downloaded on landing. The system tracks component status and recommends troubleshooting steps as needed.
Benefits depend on how fully Ahead is used. “Smart integration of Ahead predicts the best unscheduled and scheduled maintenance stops, the amount of inventory and manpower at each location and the optimum moment to exchange degraded components,” Lima says. He estimates Ahead can increase aircraft availability by up to 35%. Embraer will offer a smart-integration package to new customers and is studying structural monitoring on new aircraft.
Major MRO providers also offer predictive services.has provided data and predictive maintenance to customers for 20 years. It expects to support the before year-end.
Martin Frutiger, head of IT tools at SR Technics, says data sensitivity and security are additional hurdles to predictive maintenance by independent MROs. “Data for predictive maintenance also provides an overview of pilot performance, so it creates a number of sensitivities. It is essential that strong trust is built up between the MRO and the flight operations, engineering and technical departments of an airline to ensure all data is provided.”
Frutiger emphasizes the data differences between new and old aircraft. “New aircraft are like flying servers,” he says. More data, more parameters, more frequent data and more efficient transmission can dynamically improve scheduled maintenance on new aircraft. “If a component is performing better than expected, its replacement or overhaul can be delayed. If performance is not as good as it should be, it can be overhauled earlier.”
Fault codes and troubleshooting recommendations alone are not predictive maintenance. One major OEM has taken the plunge into a full predictive-maintenance offering.and Accenture, through a joint venture, are offering Taleris on all manufacturers' commercial airframes, engines and components. The support will be end-to-end, offering hardware for collecting and wirelessly downloading sensor data, prognostic analysis, a ground-support network, Accenture's planning and recovery tools, and transmission of recommendations to personal devices.
Raffaele Delogu, director of strategic markets for GE Avionics, says Taleris analysis will go beyond physical models to detect the unexpected and suggest actions relevant to operators. Delogu sees the major challenges as harmonizing data from multiple systems and changing airline business processes.
“OEMs will still make recommendations for repairs, but we will do advanced anomaly detection at the system and subsystem level above OEM fault codes,” explains Taleris CEO Norm Baker. “We will be able to see before anyone else the probability of degradation.”
In addition to OEMs, consultants also help carriers with predictive challenges. As head of technical operations at Finnair, Manu Skytta was having expensive challenges with his' bleed systems. Since they were flown 19 hr. a day, there was little time for maintenance and none to waste on no-fault-found removals. “We needed to do the right action at the right time,” Skytta stresses. “That way you also have materials available when you need them and you can do production planning for repairing components.”
Finnair had Airbus's Airman but had been working with Frankfurt Consulting Engineers (FCE) on optimizing assets and reducing turn-times. FCE consultants suggested applying their Anomaly Detection Software to the bleed system. “We were quite skeptical,” Skytta remembers. “They didn't know anything about the technology. All they had was parameters.”
Finnair sent flight parameters for one year. The carrier knew what had happened to the bleed system, but did not tell the consultants. Skytta just asked FCE to tell him if something went wrong and whether he had changed components. “They saw the times we had changed a component that had failed. They totally surprised us. They got the right answers,” he says.
So Finnair sent FCE data for the entire fleet for six months and did no bleed-system repairs so as not to ruin the experiment. FCE software spotted problems in the pneumatic system that Airman missed. “We changed it and the problem vanished,” says Skytta.
Often the real challenge in doing predictive maintenance, Skytta notes, is spotting real problems without too many false alerts. “The problem with OEM systems at the moment is you get too many warnings. So the troubleshooter does not trust them. And a solution is no good if no one uses it.”
Lack of data is usually not a problem: “We have thousands of parameters on Airbuses, and airlines have found ways to collect it,” Skytta notes. Carriers already use this data for flight-monitoring, as required by regulations, and some data also could be used for predictive maintenance. However, “pilot agreements might prevent sharing data for other purposes,” Skytta acknowledges. “We do not have a culture of sharing data.” That is unfortunate because the best predictions would be based on multiple-airline data.
Skytta says different data formatting is usually not a big issue, noting that Airbus manages to collect the data it needs mostly in standard format. Airman and other OEM tools can be expensive, especially for small operators, and Skytta believes consultants such as FCE might help them.
Skytta says FCE's statistical approach can look for abnormalities in any system for which there is data, including mechanical, hydro-mechanical and electro-mechanical systems, even generators of electrical power. “It would not work for electronic systems. We do not have metrics on the circuit board,” he says.
FCE has been using its Anomaly Detection Software since 2004, mostly in the power and rail industries. Aviation Project Manager Daniel Jaroszewski says the technique requires no knowledge of underlying physical processes but lots of data as it looks at the correlations among all signals from the system under study.
The aim is to minimize both false alarms and missed critical events. The software can “learn” without supervision by spotting data patterns in a mostly healthy operation of a component or system, as it did for Finnair. The longer an anomaly persists, the more likely it will turn into a critical event. If data on malfunctions is available, the software can “learn” with supervision.
Jaroszewski thinks his software outperformed Airman in spotting several bleed-valve problems because it looked at correlations of all the data, not just one signal. He says the approach is applicable to many aviation components, so long as data is plentiful, as it increasingly is with new aircraft. Apparently others agree: FCE has been talking to several aviation companies, including, in recent months.
Other consultants are also active. “No one has all the tools,” stresses Vijitha Kaduwela, CEO of Kavi Associates. He says good analytical tools are available from SAS, IBM, SAP and others. “But you cannot just buy off the shelf and go; you must customize it.”
That is what Kavi consultants, drawn fromand GE, do. Kaduwela believes there is great value to be gained in spotting chronically defective units, minimizing aircraft downtime, assessing supplier quality and finding the “bad apples” among aircraft. “There are lots of opportunities to find the outliers and fix them,” he says.
For example, text comments on job cards could help spot chronically failing units, but this data has not been digitized and mined. Text-mining techniques are now making that possible. “There is big money if you can weed chronically failing avionics items out of your supply,” Kaduwela says.