Why APM exists as an industry

So why does asset performance management exist as an industry? It's common to see statements like 80 % of all failure modes are not wear out based or time based or things like that and plugging for predictive maintenance or condition based maintenance. But let's go back into the story behind it a little bit. And first we're going to step back to the 1850s and we're going to talk about boilers. So, boilers were being fairly heavily used in coal power plants. Really, steam was being used as a main source of power, or mechanical power. And railway cars were becoming a fair bit more popular.

In all the meanwhile, in 1857, there's this club called the Polytechnic Club that was formed in the US. Now, there are a couple notable people there, but I think that the main point was to try and kind of deploy technology a bit more safely. And some of the founding members were Francis Pratt, Amos Whitney, both of Pratt and Whitney, Alicia Root, who at the time was, I think, leading Colt's firearms. And then these two people, Jeremiah Allen and Ezra Roberts. And those last two end up being kind of the first and second president of Hartford Steam Boiler, but we'll get there. So, late 1850s, this club is formed. Then, as we get into the 1860s, other events take priority. The Civil War happens. And at the end of the Civil War, there is a terrible and tragic steamboat explosion. In 1865, there was the Sultana steamboat explosion, and to date, it's the worst maritime disaster in U .S. history. So it was a horrible thing.

I think three boilers exploded almost immediately thereafter. That was the start of the company, Hartford Steam Boiler. year. Somewhat coincidentally, over in Europe, TUV was also started in the same year. So in 1866, we have both of those companies that that came into existence. Hartford steam boiler specifically was really digging into the concept of how can we operate and maintain boilers more safely. And they brought up the idea of doing more periodic inspections. Now they also were backstopping those inspections with a level of insurance. But this was really all in response to frequent and catastrophic boiler explosions. So it got to the point that in 1868, it was publicized that there was one explosion every four days in the US. So they knew this was a problem that needed to be solved. A lot of people were working on this. Fast forward from 1868 to 1880, this is actually the start of the ASME, the American Society of Mechanical engineers. And inspections are improving. Maintenance practices are improving. But really the whole underpinning of it is a lot more doing or doing more time -based inspections and time -based maintenance. In the early 1900s, there are still kind of more catastrophic boiler explosions. In Brooklyn, there was a Grover shoe factory explosion in 1905. And And ultimately manifests itself in the ASME finally coming up with a boiler and pressure vessel code in 1915. So in this first phase, it's all a lot about boilers. And really the main point was through dealing with all of those catastrophic explosions, we realized that we should do more periodic inspections and time -based maintenance. maintenance. So this was really the start of it. In parallel, it also the start of kind of insurance backstopping those inspections. But that's the first act of our story. So boilers, Hartford steam boiler, TUV, inspection and insurance. Now we're going to move a little bit farther ahead in time and talk about air travel and ultimately the start of what we now refer to as reliability -centered maintenance. So So early 1900s, 1920s, and 30s, specifically in the 1930s, commercial air travel in the United States is increasing substantially. I'm looking at the numbers. In 1930, there were about 6 ,000 passengers on commercial air flights. And then by 1934, that was up to a half a million. And between then and the end of the 1960s, they finally transitioned to the point that, at least in the US, there were more people traveling by air than by rail. So, all of that industry booms. Understandably, the FAA is focused on passenger safety. So, they're looking into this, and we're kind of right in the middle of the 1960s, and looking at overall aircraft reliability. Very focused on the engines, but just overall reliability. And it was mandated that every 8 ,000 hours, so roughly once a year, every aircraft engine needed to be overhauled. Now in looking at the reliability data on that equipment, or on the aircraft, they realized that it wasn't good enough. So there was this logic of, okay, Well, if we're doing overhauls every 8 ,000 hours, let's decrease it. Let's take it to 7 ,000 hours.

And some curious things happened. So first, some of the failure modes were happening less frequently. So that's what they were expecting. But then they got surprised because it was only a small portion of those that actually happened less frequently. Some were still happening at the same rate. So, despite inspecting and servicing things at a higher frequency, the failures were still occurring. And then the real curveball hit them in that a lot of the failure modes, or a lot of the failures, actually increased by that increase in maintenance. So that really confused and frustrated the FAA to the point that they went in and established a task force of people across industry to really try to understand what's going on here. So a consortium of people from different companies ended up looking through 12 years of historical data on airplanes. And they learned, or they effectively more explicitly quantified what I was just talking about. And what they were finding was that the vast majority of the failure modes were not actually time -based. Many of them were just random, meaning they were as equally likely to occur earlier in their life as opposed to later in their life, or some were more likely to occur earlier in their life, which has an unfortunate term called infant mortality that is attached to it. So this this study was done and after realizing this they're like okay well if there isn't a time -based component in these failure modes then it makes no sense whatsoever to do time -based maintenance on them. That brings

us to 1968 where the first spec was written for kind of maintenance strategies with these new learnings. It's historically referred to as MSG -1, and it was applied to the 747. So in that, they were looking and saying, okay, well, historically, for say the DC -8, so this is kind of pre that MSG -1, so the plane, the DC -8, when an overhaul occurred, it was on 339 components. And for every 20 ,000 hours of operation, there was about 4 million labor hours associated with it. Now we flip over to the 747, and there was a dramatic decrease in the number of items that were overhauled. But kind of the more stark point is, for the 20 ,000 hours of operation, it was just about 66 ,000 hours of labor for maintenance. So it was a reduction from 4 million to 66 ,000. The rest of the aviation industry is really excited about this and they're like, okay, MSG -1, it was essentially written for the

747. The ask by the FAA was to generalize this, so MSG -2 was published two years later in 1970, where they stripped out the 747 language and it was more generally applicable to all aircraft. So, last anecdote or example about that is the DC -10, so a more complex plane compared to the DC -8. If you remember, the DC -8 had 339 components overhauled, 4 million labor hours. Now the DC -10 just had seven items that were overhauled. So massive change and massive reduction, but an overall increase in reliability. So all the meanwhile, the Department of Defense got very interested in this and they ended up asking to write a paper on this. So that MSG 1 was published in 68, MSG 2 was in 1970. After the Department of Defense asked for United to write things up, in 1978, Stan Nolan and Howard Heap ended up publishing the first instance of reliability -centered maintenance. And just to pull one quote that I'd say is perfect in here, in some of the early pages of here, of the paper. And I quote, consequently, a maintenance policy based exclusively on some maximum operating age would, no matter what the age limit, have little or no effect on the failure rate. Now, I just want to kind of say that that last bit. So in 1978, I'm recording this in 2024. So a considerable amount of time has passed. And I feel like we are still opening up and talking about asset performance management with this point and saying, no matter what the age limit, this has little or no effect on the failure rate. So really, this is why asset performance management exists. Reliability centered maintenance as a paper was written and submitted in the late 70s. In the late 90s, I think in 1997, John Mowbray wrote the book on reliability -centered maintenance 2 .0, and it's really building on this initial work. Since then, there also has been many studies looking at submarine reliability, other kind of like naval reliability, other aircraft. And, we're consistently seeing whether it is a split of 90 % random, or not time -based, to 10 % time -based, maybe it's 20 % time -based, 80 % not. But the main point here is, because it doesn't make sense to keep track of time, or to use time as the way of driving maintenance strategy, there's the logical question of, well, then what should we do? And that sets us up for the next section where we're going to be talking about condition -based maintenance. Thanks.