| Predictive maintenance takes off at EPCOR. The Air France-KLM subsidiary specializing in overhauls and repairs of APUs (auxiliary power units) for the major commercial aircraft programs has been implementing its "Prognos for APU" digital platform for a little less than three years. The results are now very tangible, as Niels van Hofwegen, Predictive Maintenance Specialist at EPCOR explained to us in Dubai.
What are the recent developments in Prognos for APU?
A dedicated team has been set up and is continuing to develop the system. This means that we are developing software, intelligence and algorithms, while at the same time the number of APUs we are monitoring is increasing significantly. Last year, in 2019, we had a 100% increase in the number of assets, with over 950 auxiliary power units monitored by Prognos within 32 different airlines. Of course, we started with APUs from Air France, KLM and Transavia, but we now have a range of customers from around the world in the system.
Is this system very different from the one you developed for engines, Prognos for engine?
No, the system's not very different. It uses the same methods and techniques, but with different software developments. Prognos for APU concentrates on EPCOR's capabilities, so on seven different types of APU (A320, A330, 737,777 and other families) as this enables us to capitalise on our knowledge of this equipment.
For example, we have just developed an APU digital twin, as given the high quantity of data we collect and the complexity of the algorithms, we needed a different way of data visualization. For the hot part of an APU, which comprises numerous rotating parts to transfer the energy from the air to the shaft to generate electricity, our digital model enables us to extract one very important piece of data, the "turbine efficiency", which takes account of the fuel flow, internal temperatures and a host of other parameters.
This lets us see the deterioration in APU performance over time, making future decisions easier. This is unique compared with what the competition offers.
What are the other advantages of Prognos for APU for operators?
One of the advantages of Prognos for APU is that this tool has been designed inside an MRO. EPCOR is an "airline MRO" and I, like the rest of the team, am working right next to the shop. All the knowledge that we on failed APUs, are directly inserted into the system. This is something that I would define as a closed loop situation. We can monitor their APUs during operation but also during their shop visits. Finally, the only times when we are not in direct contact with the APUs are when they are in our stores. This closed loop situation means that we have full control.
Do you have any examples of the operational benefits provided by predictive maintenance?
What we were working on last year was that we predicted failures. The APU entered the shop and we could see evidence of the failures that we had anticipated with the system. For example, we had observed that a bearing was heavily damaged but still operational, whereas our predictions were for an imminent failure, which turned out to be the case. This APU operated perfectly until the last moment, just before its removal.
Our main goal is to schedule removals, which means that we schedule replacements as early as possible. Of course, our other objective is to reduce total cost of ownership, which means that the equipment needs to be removed at the right moment.
What is the next step for Prognos and EPCOR ?
The next step will be to integrate the supply chain into this closed loop too. Now that we are able to predict failures, the next step is to combine the fleet of our customers according to our shop's workload. We will then have shorter execution times (TAT), a reduction in the amount of spares and lower shop visit cost, as we are also planning to repair more parts which are currently scrapped.