CASE STUDY: Norse Atlantic Airways goes for advanced M&E analytics

Author: Sander de Bree, Founder and CEO, EXSYN Aviation

Subscribe

Sander de Bree, Founder and CEO, EXSYN Aviation Solutions introduces a cutting-edge maintenance data management system

This article is about maintenance data and M&E data analytics as applied in Norse Atlantic Airways. However, before we get into that, it’s always worth understanding the background of the business adopting any new solution. So, let’s start off with explaining a little bit about Norse Atlantic Airways

NORSE ATLANTIC AIRWAYS

With bases in Oslo, London, Miami, New York JFK and Paris Charles de Gaulle, Norse Atlantic is a low-cost long-haul airline, operating a fleet of twelve Boeing 787 aircraft, serving mainly the countries colored blue on the map in figure 1.

Figure 1

Readers might have seen the airline in the news recently, because they are the only airline in the world that has managed to land a Boeing 787 in, and take off from Antarctica. However, this was not the business challenge with which they approached EXSYN Aviation Solutions.

THE CHALLENGE

That challenge with was in their maintenance and engineering activities, as illustrated in figure 2.

Figure 2

Norse Atlantic wanted to improve performance and time management within the broader maintenance and engineering department when producing reports and having insights on the data that they have in their organization. The Boeing 787 aircraft generates significant volumes of data that can be insightful in terms of aircraft performance, enhancing aircraft availability and aircraft reliability. But as well as the data generated by the aircraft itself are the various different systems that Norse Atlantic’s engineering organization has deployed in their operations. Maintenance engineering operations hold a vast amount of actionable data that can be used to produce insights around maintenance performance, engineering performance, performance of the aircraft, etc. With multiple different systems within maintenance and engineering, you might have one system used for aircraft utilization tracking, one for maintenance program, compliance and maintenance planning, one for safety management, etc. Norse found that it was taking a lot of time and effort to get data from those multiple sources and different systems, then to collate and aggregate the data in order to get actionable insights. That was the business challenge.

Bearing that business challenge in mind, Norse decided to look into the market for a solution that would help, them to organize and make full use of all of their data and, having considered what was on offer, they identified Avilytics from EXSYN Aviation Solutions to onboard to their entire fleet and aircraft data management platform. The main reasons behind that choice are categorized in five main decision points shown in figure 3 that Norse Atlantic considered key for them to onboard or to undergo this exercise of onboarding their fleet data and their organization data on the platform.

Figure 3

First and foremost, and a very important element for the airline, is data accuracy. So, the solution had to have strong system integration capabilities with the existing systems that Norse Atlantic already had in place. The objective was not to implement another system that users needed to feed with data, but to have a system in place that was able to automatically take data from systems used in the airline today for tasks like workflow management and day to day transactions within maintenance and engineering.

Data Security is obviously important, as you can imagine. Aircraft onboard data, maintenance and engineering data; these are mission critical data for any airline. It’s data about aircraft reliability and flight safety; sometimes it’s even about competition data and some employee related data as well. Norse Atlantic specified high level and stringent data security requirements that the Avilytics platform was easily able to meet due to it being fully embedded in the Microsoft Azure Cloud.

Thirdly, one of the main key decision points was the existing functionality because, for Norse, this was basically a build versus buy decision and what they found was that, within Avilytics, they could kick start a lot of their actionable intelligence, a lot of their data insights on maintenance and engineering. That was based on the wide range of pre-built analytical functionalities in the solution that pretty much cater to the entire value chain of a typical management engineering organization; reliability management engineering, KPIs, safety management, etc.

In addition to that, and this is actually something that Norse is using heavily (see examples later), the airline is able to use the Developer function to build their own graphs, insights, and sometimes even their own machine learning models, based on the aggregated data they share into the Avilytics data warehouse. It offers the best of both worlds with a lot of pre-built functionality and, in the event that you have a particular business requirement that cannot be met, Developer enables users to build that functionality themselves.

And lastly, partner collaboration was quite an important aspect for Norse Atlantic Airways, having a flexible and agile collaboration between them as the airline, as the user of the system and, in this case, EXSYN as the as the software vendor.

Once they had selected Avilytics as the platform to onboard their 787, fleet data and for their maintenance and engineering data, there had to be a specific process of onboarding that data, which can be seen in figure 4.

Figure 4

The process started with a status quo assessment of what data is available and what needs to be available, what are the gaps and how those gaps can be bridged. Next is setting up data ingestion and, there again, automation is key: Norse Atlantic doesn’t want to have people entering the same information multiple times so does want automated data integrations between all the systems that are in use. That then activated the baseline, and, from there on, Norse Atlantic Airways continued with modeling and business embedding, to make sure that the maintenance and engineering departments are able to generate value from the data and from the insights that they are presented with. That is best illustrated by a few adoption examples that that I can take you through as an example on how Norse Atlantic Airways, with their 787 fleet, is using this kind of capability in their in their organization.

RESULTS FROM ADOPTING AVILYTICS

Here are just a few of results that have beed gained from utilizing the features and capabilities built in to Avilytics.

Reliability automation

The first result revolves around Reliability Automation, as you can see in figure 5.1.

Figure 5.1

As a Boeing 787 operator, what Norse has been able to do for well over a year, is to fully automate the creation of their regulatory reliability reporting, as well as have in depth analysis on specific reliability issues they might face in their fleet and what can be done to improve on those issues. What that means is that they have pretty much transcended from a reporting cycle, spending time and efforts to create reports, to having that time available to do actual reliability assessments and analyses, plus implement actions that positively impact overall aircraft availability, dispatch reliability and systems reliability.

Engine contract monitoring

Another quite interesting example where aircraft onboard data comes together with, in this case, maintenance engineering data coming from the from the AMOS system, is what we today call engine contract monitoring as in figure 5.2.

Figure 5.2

The example is a 787, with Rolls-Royce Trent 1000 engines which is never bought but always leased with specific leasing parameters on how lessees are allowed to utilize the engine. There are specific airports that, when you fly to those airports, you might incur utilization penalty factors such as engine derate settings that the crew is applying that can have an effect on the allowed utilization of the engines. Other factors include stage lengths, flight durations, etc. And once a user systematically exceeds their lease parameters, that can lead to significant penalty fees being imposed on the operator for over-utilizing the engine, beyond what the leasing agreement allows. Avilytics takes the engine data, derate settings, the flight operational data, stage, sector, length of flights flown, etc., and compares that to the actual engines installed on the aircraft. That gives a full overview of all the engines installed on the aircraft and how high the risk is that each particular engine might exceed the leasing agreement allowed parameters. That doesn’t necessarily mean that, as soon as there is something in red on this dashboard, that engine should be directly replaced, because that’s a costly affair, but it allows for planning in advance. It allows for making an informed decision, should this aircraft be assigned to fly that particular route or, if we do that, will it lead to penalty exceedances. By combining engine data coming from the aircraft with configuration data coming from the maintenance tracking software and flight operations data coming from the scheduling system, it allows the airline to decide how to best assign that aircraft on specific routes to avoid them exceeding allowed parameters for engine usage.

ACARS Comparisons

ACARS is another data feed coming from aircraft in the fleet into Avilytics, through the SITA network. And we use ACARS data for live data monitoring and live data comparisons, as shown in figure 5.3.

Figure 5.3

Through ACARS you have the on block and off block times which are used to calculate flight durations and automatically cross reference that with crew inputs to the flight log, in order to make sure, for instance, that any crew entry made on the flight log itself matches the actual measured flight duration coming from the ACARS system. This is a great help in making sure that the actual accumulation of flights flown, hours and cycles of the aircraft remain consistent over time and if, by any chance, an incorrect entry is made in the journey log, that automatically gets detected and corrected in the system itself. So, again, automation serves to make sure that this specific set of critical data is always updated based on actual information coming from the aircraft.

Supplier performance

Also quite nice is supplier performance as in figure 5.4.

Figure 5.4

Supplier performance is based on the data coming from Norse Atlantic’s supply chain ERP system, which allows the monitoring of turnaround times and service levels of their pool provider. As a 787 operator, the airline is covered by a pool program from a specific supplier. Avilytics allows them to monitor the performance of their pool provider, against pool agreements, against pool service levels, perhaps not necessarily, in the first instance, to enforce certain contractual implications on them, because that’s not really in the nature of the organization, but in order to see what can be done to collaborate with the pool provider in resolving some of the supply chain issues that they might face. Norse knows what kind of items they most frequently request from their pool providers. They know the turnaround times on those requests depend on the priority that they give the request, whether it is an AOG request, a normal request, an exchange in advance, etc. With that, they can take actions together with their pool provider to shorten those turnaround times and positively impacts the overall maintenance downtime of the aircraft. They might even use it to go back to their vendors to say, “You’re not compliant with our contract parameters here”.

AOG risk management

Norse Atlantic also uses Avilytics within the maintenance control center for AOG risk prediction as illustrated in figure 5.5.

Figure 5.5

Based on configuration data of the aircraft’s utilization and expected, based on data coming off the aircraft itself, the system provides an overview of risk profiles, meaning, what are current and future probabilities of each specific aircraft in the fleet running into a AOG situation due to specific failures that could occur on the aircraft, and what are the probability factors of those occurring. Those are classified as low-, medium- and high-risk events, and that information is used by the Maintenance Control Center to see what kind of risk mitigations could be taken in the event that such an AOG situation would occur on that aircraft. That’s not to say that each listed high- risk event results in a maintenance action. It is being used to think about risk mitigations in the event that that particular AOG event would happen.

Examples of such a mitigation could be, what is the next arrival airport of the aircraft that has that particular AOG risk probability if that component would fail. What would that mean, practically, if the aircraft is on that out station, or is the aircraft expected to be on home base when the AOG happens. What is the spare parts availability on the base where the aircraft is expected to be? Again, the facility is there, not necessarily to trigger a maintenance action, to preventively replace a particular component, but more to think about risk mitigations that can be taken to make sure that the aircraft gets back to the surface as fast as possible in the event that that particular AOG situation is going to occur, and yes, in the odd exception, it actually leads to a preventive replacement. We do see that that decision is mainly taken once the aircraft is actually already scheduled for a heavy maintenance event so that ground time over a multitude of days that most of the time does lead to a specific maintenance action being taken; but not always when the aircraft are in operation.

ROADMAP TO THE FUTURE

Together with Norse Atlantic, at the moment, EXSYN is also integrating further systems and data sources from maintenance and engineering as you can see in figure 6.

Figure 6

One of them is in the safety domain. SMS and safety performance indicators are important, of course, and this is where EXSYN will integrate with Norse Atlantic’s Centrik 5 Safety Management System (SMS) through native APIs in order to integrate SMS data and quality management data into the Avilytics data warehouse and provide analytical insights for Norse Atlantic’s Technical Safety Action Group to monitor safety and compliance of the airline itself.

For aircraft health monitoring, the 787, has quite elaborate aircraft health monitoring systems. Right now, the onboard data coming off the aircraft is ACARS messages and EXSYN’s flight data recorder parameters, and that is now being extended to the actual flight deck FX and maintenance messages that are recorded in the health monitoring system for the Boeing 787. These are then used to further enhance the AOG risk probability calculations that that we have seen earlier, and thirdly, the integration of electronic logbook data, both flight technical and cabin. At that point, Norse Atlantic will be using the Ultramain ELB which can be integrated with other systems so that journey log data like fuel consumption, etc., and, of course, the technical and the cabin related data from the tech logs, can be fed directly into the Avilytics data warehouse, which eventually will lead into further analysis being done on things around, journey logs, flight performance, etc.

That, of course, looks very nice and sounds cool. What we have learned from this interaction is that, if you want to adopt these kinds of advanced analytical means within your maintenance and engineering organization, it is paramount to invest in the data quality that you want to use in those dashboards. And this is also where some of the underlying integration strength sits in Avilytics, because it also helps Norse Atlantic in driving and continuously monitoring that data quality to ensure that whatever is presented to end users can actually also be trusted. You might recall that one of the primary requirements was to have high quality and trustworthy data that is being used for analysis purposes. So that remains quite an important topic throughout the entire application itself. Or as Norse likes to put it, “We have used the data several times to easily visualize how the fleet is doing, and on the typical question, what are we struggling with as an organization? They have off the bat answers, not only what are we struggling with, but also, what can we do in order to improve on that? We now have very fast and good, data driven answers.”