Yann Cambier, Senior Manager, ICF shares some thoughts on Big Data: where we are, what is changing and what it will mean

The engine OEMs have been doing predictive maintenance for a long time and have been running health monitoring for decades. MRO integrators like Air France Industries and Lufthansa Technik have also been developing solutions with, for instance, Air France Industries focusing on engine maintenance and towards predictive maintenance while Lufthansa Technik with Aviatar is working towards maintenance solutions as well as flight operations solutions. AAR, another MRO player, has developed an e-Marketplace for parts trading. But the main interest in figure 1 is to look at Airbus and Boeing. Everyone knows that airframe OEMs currently don’t have a big share of the aftermarket, but they are trying to generate more revenue in this space and, in a way, compete with the component OEMs. Through analytics, they are trying to develop maintenance solutions as well as other solutions such as flight ops solutions.

One big question of the past three years has been that, while we talk about big data and there is more data generated so it’s possible to undertake more analysis, to what ends are we doing it? In the past twelve months there has been evidence of the results of predictive maintenance. For instance, easyjet and Skywise have just signed a five year agreement and, at the time of writing, there had already been 31 instances of Skywise correctly predicting faults before they occurred in service, allowing the carrier to intervene and remove components before they failed, thus avoiding 31 delays and/or cancellations. Cathay Pacific has tried Honeywell’s predictive maintenance trial program and has reduced its APU related delays by 51%. Delta has hired a team of data scientists to work on predictive maintenance for its fleet and has significantly reduced engine events as a result. They avoided 1,000 engine events over one year, achieved a 100% completion factor for 241 days during 2017 with a 98% reduction in maintenance related cancellations over the period 2010 to 2016. That suggests that, for airlines, one objective of predictive maintenance is to improve dispatch reliability.

In this article we’ll cover three areas of interest. We’ll review the current industry situation; then we’ll look at the latest market developments; and finally we’ll consider some hypotheses on what are the implications of this digital race for different players in the market.

Looking at fleets, there are currently about 29,000 aircraft in service (figure 2).

That figure is predicted to grow to 39,000 aircraft by 2027. But the main point of figure 2 is to show that the next generation aircraft (Airbus A320NEO and A350, and the Boeing 737MAX and 787) while currently only a very small proportion of the fleet, are expected to be about 50% of the fleet in 10 years’ time.

But, as we said above, the big question is that, although there will be a lot more data, what can be done with it, how can value be derived from it and, for airlines, how will that benefit their operations? For MRO suppliers it’s how can they use that data to generate more revenue or reduce costs?

Putting some of this into context (figure 3), aircraft operations can be divided into nine phases from flight planning to post-flight debrief and then maintenance.

Looking where the costs are incurred, the main areas are taxi, enroute, approach and maintenance. The first three of those phases are heavily linked to fuel consumption. At the bottom of the chart, it can be seen that most of the operating costs are linked to fuel, employee costs and then maintenance. But then, if we look at these phases in terms of ‘what is the digitalization of each of these phases?’ (figure 4) we see that maintenance is one of the least digitized phases of aircraft operation.

Most airlines already have a flight planning system in place from say Lufthansa Systems with Lido or the Jeppesen suite. A lot of airlines also have electronic flight bags (EFB). However, when it comes to maintenance, there is a lot of work going on, a lot of solutions being developed and tested but nothing much has changed over the past twenty years in terms of the way people do maintenance and how they maintain their records, etc.

Today, the big topic is predictive maintenance. As we already know, engine health monitoring has been happening for a long time, at least twenty years. Now, there is an effort to try to do the same with other aircraft systems. As figure 5 shows, there are a few players using aircraft data: the OEMs who want to make their systems more reliable; the Maintenance Control Center which has to do it from a regulatory standpoint; and then there is the Engineering and Maintenance division who have to try to keep the aircraft in the air. The more airlines can predict the serviceability of their aircraft, the easier it will be to avoid technical delays and cancellations.

Looking at the data value chain, it runs from acquisition with these new generation aircraft mentioned above, transmission which can be ACARS or a technician using a USB stick… all the way to undertaking the actual maintenance action and generating the records of that maintenance activity (figure 6).

If we think of aircraft health monitoring, it’s very much a function of analysis and MRO action. With analysis, there are two options. First with the typical way of diagnosis: is the part serviceable, yes or no? If yes, great; if no then it will have to be removed. The other type of analysis is prognosis: is the part serviceable, yes or no, and if yes, for how much longer will that be true? It might be serviceable today but might not have much life remaining and, therefore, it could still be sensible to remove it. Coming to the association with the MRO action, if the part is serviceable the decision might be to leave it on the aircraft or might be to remove it because it can be expected to fail in a short time which has the potential to become an aircraft on ground (AOG) delay and/or an unscheduled maintenance event.

That’s the context, but what are MRO Players doing with regards to digitalization? In the last three years, there have been a lot of partnerships established across the value chain (figure 7).

For instance, Rolls-Royce is partnering with SITAONAIR and that partnership centers around data cleaning and data transmission. Boeing is working with Microsoft from the Cloud storage standpoint. Airbus has partnered with Palantir which is more about the analytics, and then Lufthansa Technics with FLYdocs which is about record keeping. Looking at the main players in the industry (figure 8) we can see how all of them are taking slightly different approaches. Airbus has a lot of partnerships in place such as with Rockwell Collins for data acquisition, with transatel for connectivity on the ground and transmitting data over 3G and 4G, and Palantir on analytics. Whereas, looking at Air France Industries, they have decided to do a lot more in-house on the grounds that the main business is an airline, they know their aircraft and how they perform plus they have maintenance expertise making their own business the best place to develop their own solutions.

Now, looking at product lifecycle, the innovation stage is now passed (figure 8); people have been developing maintenance solutions for a long time. Now we are more in the early adoption phase.

Increasing numbers of airlines are adopting these processes but it’s not yet a wide majority. And, if we look at the product: previously, there were products that addressed one key issue. Airman was the innovative predictive maintenance solution that was focused very much on maintenance. Then, in recent years, there have been developments such as Safran with their SFCO2® which focuses on fuel consumption while GE FlightPulse™ focuses on flight operation. Now we are seeing the likes of Airbus with skywise and Lufthansa Techniks with AVIATAR trying to achieve most of the things that are possible but through one platform: maintenance related solutions, flight ops, some of them do training; they are trying to cast a wide net and do everything under one umbrella.

To focus on one case, Airbus with skywise, they claim to have around two thousand aircraft participating in their platform (figure 9); so there is definitely an industry interest in the program.

It can be seen with easyJet, Air Asia (see more below) and others that there are airlines willing to participate and work with Airbus to develop solutions. However, the big issue is that, for Airbus to have a platform that is industry relevant rather than just airline relevant, they need to cover a large proportion of their fleet.

If today there are about ten thousand Airbus aircraft operating around the world, let’s assume that they need at least 50% of that fleet participating in skywise for it to be regarded as an industry-wide solution. If there are between 1,700 and 2,000 aircraft currently in skywise, from where will they get remaining three thousand aircraft to get to that industry-wide solution?

So what we did at ICF is we split the fleet into mid-size (less than 50 aircraft) and large airlines (50 or more aircraft) (figure 10) and the assumption is that, because we are still in the early adoption stage, small airlines are less likely to have enough data samples to be relevant for skywise and are less likely to be drivers behind skywise development so we put them as last in our considerations.

In the mid-size to large fleets, there are about five thousand aircraft but, if we look at them, half of them are not aligned with any specific maintenance providers whilst the rest are with Lufthansa Technik and Air France Industries. The conclusion is that, if Airbus wants to have an industry-wide solution, and that meant having around 50% of the fleet part of skywise, they have to convince most of the non-MRO aligned airlines to work with them on skywise or they need to convince airlines that work with Air France Industries or Lufthansa Technik on the component MRO side to work with them on the digital maintenance platform.

Whilst that might seem like a challenge, why would someone do their component PBH (Power By the Hour) with Lufthansa Technik but at the same time, work with Airbus on predictive maintenance, there are examples of people who are considering that. Air Asia has been a long-time client of Air France Industries under component PBH and has chosen to participate in skywise with Airbus and is also working with GE Aviation on FlightPulse and FOQA – Flight Ops Quality Assurance and Flight Operations performance.

So, that’s what’s happening in the market: there is a growing fleet with increasing numbers of connected aircraft joining the fleet in the next ten years. There are a lot of suppliers developing different solutions. Now we need to consider what the implications of all this activity might be.

First, what is the benefit for the industry? We’ve tried to model what could be the benefits for airlines from a cost saving viewpoint. For an airline, digital is a unique way of generating new opportunities for passenger experience and ancillary revenues. It’s also a cost reduction tool, reducing maintenance costs and fuel costs.

The estimate is that predictive maintenance will improve technical dispatch reliability, it will drive a reduction in ‘no fault found’, and will support a reduced inventory and improve labor productivity. That could generate about $3 billion of savings for the industry. Then, on fuel costs, with tools such as live weather updates that enable pilots to avoid storms, also better speed and altitude optimization… it could all generate about $1.7 billion of savings. And finally, digital solutions can be used to improve the turnaround phase. Each delay minute has an associated cost of about $80-100, so if turnaround delays can be reduced, that will save money, according to our calculations, to the tune of about $0.8 billion across the industry.

On the maintenance side, the question is, ‘if you’re a supplier, to get more revenue from airlines, do you use digital solutions to generate cost reductions and try to sell a spend reduction of, say, $5 million through lower delays and lower maintenance costs’ or ‘do you use digital solutions to be better at maintenance, to know better your product and therefore be more aggressive with your contract and win more contracts?’

Taking that question and looking at the component PBH for a fleet of 50 or 60 narrow-body aircraft, it would have a value of $5m to $8m per annum: typically, a component PBH is a five year contract; so with a full cost of between $25m and $40m, that will vary depending on the age of the fleet, the scope of the contract and so on. On one side of the equation, digital can be used to win $25m to $40m contracts or, on the other side of the equation, going to individual airlines and telling them that predictive maintenance is going to reduce their technical delays and their technical cancellations. If you think that delays cost $80-100 per minute and a cancellation can cost between $25,000 for a narrow-body aircraft and $110,000 for a wide-body aircraft, that’s a lot of delay and cancellation costs to avoid to generate similar levels of revenues.

There are opportunities on both sides but our view at ICF is that, if you are an MRO provider, an OEM or an MRO integrator and you have limited bandwidth, then your focus should be more on using digital to be more competitive in the marketplace and therefore bid more aggressively for component PBH programs rather than trying to sell your digital solution on the basis that, if the airline uses it they will reduce their technical delays; because the revenue opportunity is much lower.

The first one sounds very basic but, do you have all of your reports in a database? We’ve had quite a lot of discussion with component OEMs, for example, and some of the feedback was really surprising: a Chief Information Officer telling us that they have functional analytics tools and they have data scientists but the shop findings reports are still on paper. That means that you could do great analyses but there is no data to analyze. So, the priority is to get your infrastructure in place so that all processes and all records are actually in a database to be exploited.

The next question relates to the fact that 20 to 30 years ago maintenance used to be a core activity for airlines whereas nowadays most airlines outsource it and try to think as little as possible about maintenance: they just want it to cost the least possible. So will we see the same happening with other activities that airlines currently do? For instance, easyJet has outsourced all of its supply chain to AJW Group which means that they don’t do the maintenance and they now don’t do the management of the maintenance anymore. Will the same happen with, say, flight operations and the Operating Control Center (OCC)? Will we see airlines going to Boeing and to Airbus to say ‘here is our current performance through the OCC; if you can guarantee that you will perform at the same level, you can undertake that as an outsourced function. Airlines can then focus on their revenue opportunities and leave the cost side to sub-contractors.

The third point is about the interaction between the players. Before, the airframe OEMs made money from selling aircraft; the component OEMs from supplying Airbus and Boeing on very low margins and making their money on the aftermarket. But now that everyone is fighting for the aftermarket, the relationship has changed, component OEMs used to be suppliers to Boeing and Airbus but now they’re competing on the aftermarket as well, so that changes the dynamic. Then there are players like Amazon, who used to be a cloud service provider but are now a cargo carrier based on the fact that they can manage every step of the logistics very well. So can they do the same with an airline?

There is a lot of change in the marketplace so the final question is, ‘who will win, who will lose?’

Mergers and acquisitions (M&A) activity used to focus on R&O (repair and overhaul ) with OEMs acquiring shops with specific repair capabilities, now the focus seems to be on acquiring digital capabilities. Where M&A is not possible, OEMs are focusing on large-scale partnerships. For instance, while Boeing cannot acquire Microsoft, they can be partners with them as Airbus has similarly done with Palantir. And then, because of the change of dynamics, people who are very good at data analytics but know very little about maintenance, are now playing in the market and competing with the likes of Lufthansa Technik and Air France Industries in that market space.

Let me close with the summary that digitalization offers airlines a potential for efficiency that is only just being understood. Suppliers need to create the right partnerships and value propositions. The time to act is now.

 Yann Cambier is a Senior Manager with ICF. Yann has worked both in industry and as a consultant. He was part of the executive team at Airinmar (an AAR Corp. company), where he led the digital strategy development, as well as the strategic sourcing of maintenance contracts for AAR’s PBH programs. As a consultant, Yann has worked on operational projects, strategy development projects and due diligence projects. He supports airline clients, aerospace suppliers and private equity firms.

ICF is a global consulting services company with over 5,000 specialized experts, but they are not typical consultants. At ICF, business analysts and policy specialists work together with digital strategists, data scientists and creatives. They combine unmatched industry expertise with cutting-edge engagement capabilities to help organizations solve their most complex challenges. Since 1969, public and private sector clients have worked with ICF to navigate change and shape the future.