00:00:04 Intro to aircraft market’s unique challenges.
00:02:01 Joannes on aerospace maintenance’s crucial, complex role.
00:03:56 Supply chain complexities in aerospace discussed.
00:06:00 Defining “repairable” in aerospace context.
00:07:44 Start of talk on maintenance companies’ stocks.
00:09:43 AOG incidents, impact of missing parts explained.
00:11:30 Collaboration among airlines in industry described.
00:12:19 Transition to forecasting in airline industry.
00:14:10 Complexity of aerospace supply chain examined.
00:16:01 Challenges of aircraft parts repair discussed.
00:16:56 Complexity of forecasting demand explained.
00:17:35 Probabilistic forecasts in parts maintenance insight.
00:19:24 Cost of large parts stock discussed.
00:21:20 Forecasting experiences for aerospace clients.
00:23:56 Parts exchange complications in inventory.
00:26:33 Minimizing AOG incidents per dollar, introduced.
00:28:00 Aerospace investment strategy explained.
00:28:30 Need for probabilistic approach.
Kieran Chandler and Lokad’s founder, Joannes Vermorel, discussed the intricacies of the aerospace industry. They highlighted the costly groundings, on-time aircraft movement necessity, and the paramount importance of safety. Vermorel emphasized the unique supply chain due to the industry’s maintenance needs and the lengthy lifespan of airplanes. Chandler compared an aircraft to Lego blocks, pointing out the scale of operations in maintenance. Vermorel clarified that a “component change” often implies inspection rather than repair. He also discussed the financial implications of maintaining an aircraft, the industry’s collaborative nature, the criticality of forecasting, and challenges in counting parts due to the supply chain’s dynamism.
Kieran Chandler, the host of the interview, began by highlighting the complexities and critical aspects of aviation safety and profitability. He pointed out how grounded airplanes can cost airlines up to $200,000 per day, making on-time aircraft movement crucial.
In response, Joannes Vermorel, founder of Lokad, described the uniqueness of the aerospace industry. Safety, he pointed out, is paramount in this sector to an extent that failure can lead to immediate loss of life. The industry must handle a diverse range of technologies, including high-pressure and high-temperature parts, electronics, lightweight materials, and components designed to withstand varying weather conditions. Despite these complexities, affordability remains a key consideration as air travel has become increasingly accessible.
Vermorel further explained that an airplane’s lifespan can span several decades, necessitating meticulous maintenance. Unlike other products, an aircraft isn’t simply utilized until its retirement. Instead, virtually every part of the plane undergoes regular inspection and possible repair. This maintenance process is what keeps aircraft in the air, and the aerospace supply chain is completely oriented towards this goal, adhering to exceptionally high safety standards.
Comparing an aircraft to a set of Lego blocks, Chandler inquired about the scale of operations involved in aircraft maintenance. Vermorel replied that a commercial aircraft comprises about two or three hundred thousand different parts. Although many of these parts may only be changed infrequently or not at all during an aircraft’s lifetime, some components, like tires, need to be replaced after a certain number of flights.
The discussion then turned to the complexity of the aerospace supply chain. Vermorel described how, unlike conventional supply chains that move forward from producers to consumers, the aerospace supply chain operates differently due to the industry’s unique maintenance needs. He referred to the concept of “component change,” which involves regularly unmounting parts for inspection and repair due to the long lifespan of aircraft.
Vermorel’s explanations throughout the conversation emphasized the intricate balance the aerospace industry must maintain between safety, technological diversity, cost-effectiveness, and meticulous maintenance. This balance, coupled with the immense number of aircraft components and the unique nature of the aerospace supply chain, serves to illustrate the complexity of the industry.
Vermorel first describes the unique nature of the aerospace supply chain. Unlike traditional supply chains that are linear – moving from producer to consumer – the aerospace supply chain consists of a series of loops. Aircraft components are produced, utilized, then removed for repair when deemed unserviceable. These components are sent to repair shops, where they are restored to serviceable condition and then reinstalled on aircraft. This process repeats for the lifespan of the components.
Vermorel goes on to explain that while this system may seem unsettling to outsiders, it is in fact a safety measure. The vast majority of parts on an aircraft (approximately 90% by value) are repairable. However, “repair” in this context doesn’t necessarily imply that a component is broken; instead, it often simply means that according to its maintenance schedule, the component needs to be inspected and validated to ensure it’s still fully flight-capable.
The guest then emphasizes the financial implications of maintaining an aircraft. Keeping an aircraft in the air is a priority because if it’s not flying, it’s not generating income. This financial pressure, coupled with the goal of maintaining safety, can lead to an “aircraft-on-ground” (AOG) incident if just one part is missing or unserviceable. In such situations, airlines need to source a serviceable replacement part as quickly as possible, often at a steep cost.
Vermorel also discusses the role of collaboration within the aerospace industry. Despite the competition between airlines, there is also a high degree of collaboration. This collaboration is necessary due to shared infrastructure and common security interests. Moreover, when one airline experiences a flight cancellation, passengers often have to book with another airline, further fostering this collaborative relationship.
Finally, the conversation shifts to the topic of forecasting in this highly complex environment. Vermorel points out the importance of forecasting not only demand but also all the interactions and occurrences within the aerospace supply chain loops. The objective is to ensure that no piece of equipment stays idle for too long. If an extra piece of equipment is bought and not mounted on an aircraft, it will wait until requested. However, if there are other compatible pieces of equipment, they can also be used and repaired.
Vermorel highlights the uniqueness of the aerospace industry, where every component removed from an aircraft is replaced with another, creating a dynamic cycle of mounting, dismounting, and repair.
Vermorel introduces the concept of instant repair, an ideal scenario where dismounted parts are immediately repaired and ready for remounting. However, reality deviates significantly from this ideal due to various factors that delay repair: returning the part, transporting it to the repair shop, the workload and capabilities of the repair shop, and the nature of the required repairs.
A significant challenge is forecasting not just demand, but also these delays, which directly influence the number of spare parts needed. Demand is erratic, especially for rare parts needed only a few times a year. Despite some advances in predictive maintenance using sensors, the process remains unpredictable. Nonetheless, Vermorel notes there is some structure to this unpredictability, depending on maintenance frequencies and the aircraft’s needs.
The conversation shifts to the cost implications of keeping large amounts of inventory. While it may seem sensible for companies with abundant resources to maintain large stocks to prevent aircraft groundings, this approach has limitations. Aerospace parts are long-lived but may become obsolete as aircraft technology advances, potentially resulting in substantial inventory write-offs.
When asked about the challenges he faces in forecasting for aerospace clients, Vermorel brings up the unexpected difficulty of simply counting parts. The dynamic nature of the aerospace supply chain, with parts constantly in transit for repair, replacement, or even lending to competitors, complicates accounting. Another complication arises from the fact that parts are often made up of sub-components, which can be exchanged individually. A primary challenge in aerospace is having a clear picture of non-aircraft-attached inventory to identify where investments should be made.
Kieran Chandler: So Joannes, perhaps we can start with just a real basic overview of the aircraft market and what is it that sets this industry apart?
Joannes Vermorel: Aerospace is pretty unique due to a series of factors. First, safety is paramount to a degree that if you fail, you literally kill people swiftly. There are other industries where security is paramount, but I would say aerospace has to deal with a diversity of technologies that is absolutely huge. A chemical factory is very dangerous and you need to pay a lot of attention to security, but in an aircraft, you have everything. You have high-pressure, high-temperature parts, auto parts in a vacuum, and parts that are super light yet highly durable against diverse weather conditions. And all in all, you also have to be very cost-effective because nowadays you can travel very far for amounts of money that are actually extremely low and that are only a small multiple of the cost the company has to invest.
Kieran Chandler: Let’s talk about the maintenance. How does that work in practice?
Joannes Vermorel: An aircraft is going to be active for about three or four decades, and it requires active maintenance. It’s not like you’re buying an aircraft, and it just flies until it’s retired. Everything in an aircraft is going to be inspected and potentially repaired. Maintenance is literally what keeps aircraft flying. So that means the whole aerospace supply chains are completely geared toward keeping the aircraft flying safely by maintaining them according to very high standards of security.
Kieran Chandler: If we take a Lego brick example, when we talk about an aircraft, it’s basically made up of many blocks. How many components are we talking about here? What’s the scale of operations?
Joannes Vermorel: A commercial aircraft is composed of about two or three hundred thousand different parts. Many of these parts will be changed very infrequently or may not even be changed once during the entire life of the aircraft. But some of them are changed quite frequently, like the tires, for example, which need to be changed after a certain number of flights.
Kieran Chandler: With so many hundreds of thousands of parts, it’s going to have a really complex kind of supply chain attached to it. What are the complexities that we really associate with this supply chain in aerospace?
Joannes Vermorel: Unlike most supply chains, which are just forward supply chains going from producers of raw materials to end consumers, in aerospace, it’s completely different. It’s mostly about the fact that whenever you mount something on an aircraft, you unmount something else. This is typically called a component change. You need to do that because aircraft are very long-lived, but all their parts are inspected on a very regular basis. So, you spend your time unmounting parts, repairing them, and inspecting them. You spend your time dismounting and remounting components, and having those components repaired. Once they’re ready to fly, they’re deemed serviceable and can be serviced on an aircraft. Your supply chain is essentially a series of loops where components are taken off aircraft, now unserviceable, moved toward repair shops until they’re repaired. Then the components are deemed serviceable and they’re mounted back to the aircraft. So, your supply chain isn’t something like a forward producer to consumer. It’s a cycle between aircraft and repair shops.
Kieran Chandler: I’m not the most comfortable flyer. However, this idea of having a large proportion of parts being repairable is a bit of a funny idea. What percentage of the parts on the plane are actually repairable and how many are fully replaced by new parts?
Joannes Vermorel: I don’t know the exact percentage, but in terms of value, probably something like 90% of the value of the aircraft is made of expensive components, called rotables, that can be repaired. The vast majority of the components can be repaired because the safety standards are so high in aerospace. A repair doesn’t mean that the part is broken. It’s more like you have a component maintenance manual that defines a maintenance schedule. When you say that a part has to be repaired, it’s not that it was broken. According to its maintenance manual, the part has to be taken off the airplane to be inspected and validated that the part is still fully flight capable. Usually for many parts, nothing really happens during the repair. It might just be an inspection, and then the part is deemed perfectly safe and remounted on the aircraft. Some parts are actually repaired, like the blades in a jet engine or the tires used to land the aircraft.
Kieran Chandler: Let’s talk about the stock the maintenance companies have to hold. The idea behind the plane is it’s now going to spend probably more time in the air than it does on the tarmac. Does this mean the airline companies are having to hold huge amounts of stock?
Joannes Vermorel: Yes, the key idea is that when you have an aircraft nowadays, you want to keep them flying all the time. An aircraft is a very expensive piece of equipment and if it doesn’t fly, you’re losing money by the hour. You could actually use this aircraft to transport passengers. The whole idea behind low-cost airlines was to have aircraft that fly literally all the time and for every single trip they make, they’re full of passengers. The economics behind low-cost airlines isn’t about cheaper maintenance; the maintenance is exactly the same. It’s defined by the original equipment manufacturer or the aircraft manufacturer, so you can’t adjust anything. You just want to have your aircraft fully utilized. But in terms of maintenance, you have a very specific asymmetry. If you have only one part that is missing, this part, no matter how cheap, can keep your aircraft on the ground. That’s called an AOG incident, Aircraft On the Ground. It just means that something is missing to properly maintain the aircraft.
Kieran Chandler: What options are open to an airline company if they’ve got this one component missing?
Joannes Vermorel: Worst case scenario, they need a replacement aircraft, which is very expensive and it’s probably going to get even more expensive in the future.
Kieran Chandler: Because, as we see, all companies are becoming leaner and leaner. It means that no company has any substantial interest in maintaining surplus capacities for flight capacity. So, your alternative is to figure out how to get a spare part. When you have a part that is missing, what do you do? You want to get a spare, and it needs to be a serviceable one, so that you’re allowed to mount it on your aircraft. The distinction lies between unusable and serviceable parts.
Joannes Vermorel: That’s correct. Most airlines and aerospace companies have Aircraft on Ground (AOG) services. These are like desks where you can actually call and request an emergency quote to say, “I need this part, can you provide it to me?” The answer can be “yes,” but often at a steep price.
Kieran Chandler: And then, the airline companies care about this in very short timeframes, right? I mean, in the order of magnitude of one hour, just to have the parts delivered to the airport where it’s needed?
Joannes Vermorel: Yes, typically if you want a part delivered in one hour, it means that it has to be a part that is coming from another airline from the same airport. Otherwise, it’s going to take more time than that.
Kieran Chandler: So what you’re saying is the airlines are actually working together quite a lot. Isn’t it in their interest to see another company lose money and another company sort of fail?
Joannes Vermorel: Well, in the long term, yes, but in the short term, the aerospace industry is highly collaborative. Something that people may not realize is that all these companies compete, but they collaborate a lot. If your flight is canceled, chances are that you’re going to take the flight with another company. All these companies, they collaborate for security and share the same infrastructure at the airport. So yes, there is a lot of competition, but there is a lot of cooperation as well.
Kieran Chandler: Okay, let’s sort of move on now and talk about the forecasting side of things. From what you’ve described, we’re talking about hundreds of thousands of components. You’ve got repairable loops, potential collaboration between different companies. It sounds like a complete mess if we’re being honest. So, how can we even begin to approach forecasting for this sort of scenario?
Joannes Vermorel: That’s very interesting. First, you have to understand that you typically want to forecast not only the demand part, but pretty much all the interactions and all the things that happen in your aerospace supply chain loops. When you buy a piece of equipment, you’re going to be stuck with it pretty much forever. Remember that this piece of equipment is repairable. If you buy one extra piece of equipment and it’s not mounted on an aircraft, it’s just going to wait until it’s actually being requested.
But the thing is that if you have other compatible pieces of equipment, those pieces of equipment may be used as well, and they get repaired. So, if we compare to a traditional supply chain situation where you are thinking in terms of replenishments, you have clients, they place orders that consume the inventory that you have, and when your inventory is a bit too low, you place a reorder.
Kieran Chandler: And this replenishment is performed based on a demand forecast, but how does this situation translate in aerospace?
Joannes Vermorel: An aerospace situation is very different because if you have a part that is requested from you, it means that this part is going to be serviced to the end consumer,
Kieran Chandler: Which is going to be an aircraft at the end of the day, mounted on the aircraft. And if you mount anything on an aircraft, you dismount something. You know, it’s about the preservation of mass of the aircraft. So if you mount something, you dismount something. Therefore, there is a part that is going back to you.
Joannes Vermorel: One tricky aspect is that if you had instant repair, you would only need one spare part of every kind for your entire fleet. Whenever you dismount something, you immediately repair that part and this part becomes serviceable again and can be mounted on another aircraft immediately. So, you would only need a tiny amount of inventory of parts because you would just instantaneously repair the part. Therefore, whenever there is a demand, you just ship the free part that you have, get back the other one, repair, and be available to service the part again. But the reality is that those repairs take time. I mean, there are plenty of things that take time. First, the end consumer will have to return the part which may take an indefinite amount of time, so that there is some variability. Then it has to be transported to the repair shop again, which is also time-consuming. The repair shop needs time to do the repair and the amount of time it takes to do the repair depends on the workload that the repair shop has. The problem is that the workload depends on whatever other parts they are repairing.
And then, once it’s repaired, the time also depends on the type of operation that has to be done on the part. Imagine a repair shop gets a big component, but they only have capabilities to do certain classes of maintenance operations. They open the component and they decide, “Oh, this type of repair, we can’t do it. We don’t have the accreditation. We don’t have the equipment to do this maintenance operation.” So sometimes a repair shop will get a component, open it, and decide that they cannot do the repair, and they have to forward the component to another more specialized repair shop somewhere else to do that, and then get back the component.
So, it’s not just about the forecasting questions or it’s not just a long-winded answer about forecasting the demand. It’s forecasting all those delays and ultimately, how many extra parts that you have that are not readily attached to any aircraft that you’re actually servicing.
Kieran Chandler: I can really understand how you can forecast the demand. I can understand how a certain set of plane tires will do a certain amount of takeoffs and landings before they need to be replaced. But we’ve got so much variability with this kind of repairable process. How can you possibly approach that? How can you possibly know if something is going to be repaired in two weeks or two months?
Joannes Vermorel: The short answer is, you don’t know, and that’s the trick behind probabilistic forecasts. There is an irreducible uncertainty about when you will need a specific part. Especially when you have, let’s say, a sizable company with a fleet of a hundred aircraft, you have some parts that you will only need like three times a year for your entire fleet. So, it’s very rare, it’s going to be very erratic. Although there are some people who are making progress with predictive maintenance which involves having sensors on the components to predict when they will have to be maintained, overall, it still remains a very erratic process. But the good news is that there is some structure to this variability
Kieran Chandler: It’s not completely random and you have this one big asymmetry, which is, no matter which part is missing in the end, it will keep the aircraft on the ground. I mean, if it’s a part that is called a no-go, basically if you don’t have this part, your aircraft’s on the ground. For any no-go part, this part has a potential to keep your aircraft on the ground. So, it’s not just the average demand that is of interest. What you really want to optimize is the best supply chain decision, which is to have an extra part in stock or not, considering the cost of the part and the cost of not having the part.
Joannes Vermorel: And this cost is a very interesting concept because these airline companies are multi-million-pound companies. They’ve got a lot of money behind them and the cost of having an airplane just sat there on the tarmac is incredibly large.
Kieran Chandler: So why can’t they just have lots of parts in stock and take that sort of cost within?
Joannes Vermorel: I mean, they already do. For our largest clients, we are talking of stocks that are valued above 1 billion euro or dollars. The idea of just having an enormous amount of stocks has basically been tried. But there are limits. You see, if you have literal mountains of stocks, those parts are very long-lived, so it’s an asset that you can value in your company books. But ultimately, if you have a mountain of parts attached to an aircraft or type of aircraft that stopped flying, or your end clients don’t have those types of aircraft anymore, then you end up with a massive inventory write-off. You can’t just say, “I’m going to have mountains of inventory and they will preserve their value forever.” It’s not entirely true. Aircraft are very long-lived, but in a given market, customers expect to have certain types of aircraft that are less noisy, that pollute less and whatnot. So, you cannot assume that the parts you’ve acquired will preserve their value forever. And again, airline companies now are trying to be leaner than ever.
Kieran Chandler: Okay, and I think it’d be really interesting for our viewers to sort of hear about some of your experiences. What are the real challenges you’ve faced when it comes to forecasting for aerospace clients?
Joannes Vermorel: One of the challenges that is very surprising is just counting the parts. It might be surprising because, in aerospace, you have complete traceability. For any rotatable part, you typically have access to the entire history of the parts, where it was maintained, the aircraft where it was attached during its life, the number of flight hours and flight cycles it has lived through, the maintenance program for this component, so you know everything. So why is just counting parts so difficult? The answer is that in aerospace, everything is kind of moving all the time.
Kieran Chandler: So you want to know how many parts do you have?
Joannes Vermorel: Yes, one of the KPIs you want to count is for every type of part, you want to know how many parts you have that are not directly attached to your fleet, that are free to be not lying like not attached to an aircraft, and that can be repaired and that can basically build up your stock of serviceable parts. But it’s very tricky because you have parts that are all over the place all the time. You have parts that are being shipped, on their way to an aircraft. You have parts that are on their way back from an aircraft after being dismounted.
You know, unserviceable parts are also back from an aircraft but still serviceable. Why? Because in a big maintenance operation, time is crucial. The crews operating the maintenance operation ask for more parts than they actually need due to uncertainty. So, you end up with serviceable parts traveling both ways, unserviceable parts traveling back, and parts being repaired with varying degrees of delay and complication.
Plus, let’s not forget about the AOG desk I mentioned earlier. You can serve your competitors by giving them access to serviceable parts, but it complicates your calculations. You may be doing a standard exchange or lending parts to the rest of the ecosystem, and they do the same to you. So, if you say you have three serviceable parts, chances are you’ve actually landed one but borrowed one from a competitor. In the end, it becomes a swap where you send a part and receive a part. This gets complicated because components are made up of many subcomponents. You can end up with exchanged components, further complicating the situation. The main challenge in aerospace is having a clear picture of your inventories that are not attached to your aircraft, so you know where you should invest. Accounting can be quite difficult. It’s not just about knowing what serviceable parts you have right now; it’s about understanding the overall state of your parts across all your networks, so you know where to invest.
Kieran Chandler: I see. And you mentioned a really interesting word there: KPI. From an optimization perspective, what is it that you’re trying to optimize? Are you trying to optimize profits for airlines or reduce delays?
Joannes Vermorel: It varies, but one of the main KPIs for aerospace is how many AOG incidents, or aircraft-on-ground incidents, can you prevent per dollar invested. The idea is to own inventory and maintain your fleet properly to keep it flying all the time. The performance metric of interest is how low you can go in terms of dollar AOG incidents. It’s easy to spend millions of dollars per AOG incident to eliminate them completely, but the challenge is to shrink that number significantly. This KPI helps drive your investment decisions, focusing on the parts that will give you the highest improvement in terms of AOG incidents avoided.
Kieran Chandler: So, the key takeaway is not to apply classical supply chain models like safety stock to aerospace. It won’t work. Instead, you need a probabilistic approach that embraces the fact that your supply chain is built on loops rather than a forward flow from producer to end consumer.
Joannes Vermorel: Exactly. Don’t try to apply traditional supply chain models to the aerospace industry.
Kieran Chandler: Well, that’s all for this week. Thank you for sharing your insights into the aerospace industry. It has been really interesting. We’ll be back next week with another episode. Thanks for watching.