 Right, okay, I have 10 minutes to tell you this story. Actually, and maybe a second story as well. The first one is this delightful rabbit hole that I fell down about, airline noise and acoustics and mapping and stuff. But the second one, Karen, during our keynote this morning, mentioned that this group of technologists, us, we have a unique ability to go and explore questions and topics which have some social value, something about the information of the world around us, and then present them back in an intuitive manner. And that's kind of the second underlying story, but I'm getting ahead of myself. So, hi, I'm Ewan. You'll find me here, I work as a developer and the developer relations team at SparkPost. I am a self-taught developer and have been working in that job career-wide, but I am not, and this is possibly more important, not a mathematician trained, not a physicist and not a pilot. So, if the people in the audience who have those professions or have those skills, I apologize if I mangle your profession at any point, but I'll roll on nonetheless. So, back to the story. This is the Clyde Valley in Scotland, it's where I live. That's the city of Glasgow in the middle that you can see. The bottom pin there is my home, my family and I live there, but we're looking at moving somewhere out of town and so that really chunky set of hills up at the top of the map there, that's called the Campsie, the Campsie fells, and the pin at the top there is the town that we were looking at a moving to. And as we were exploring it, we realized there was a certain amount of airline traffic overhead and so the question was raised, is that gonna be a problem if we move there? Is it gonna be too loud to live there? And so, usually I dismiss questions like that because it's this huge rabbit hole. I'm thinking I don't have the skills or the knowledge or the formal training to explore and answer that question to my own satisfaction, but this time it seemed like a relatively tractable thing to do, so I jumped in and I broke it into pieces like I would any development task that I usually take. That doesn't mean I went to Stack Overflow and typed in this question because it wouldn't have worked. However, there are actually answers to this question. There are maps that are provided in the US and in the UK for this. However, they're built with a regulatory compliance in mind rather than any objective answer which I was looking for and some of them don't work particularly well and they're built by organizations with a different set of requirements than I had. And plus, if I just use those, I wouldn't have learned anything, right? So I went off and explored. So the first question I had to ask, the first sub question was, where are all these planes? And interestingly, Jason covered a part of that in his talk earlier, which is a nice crossover. So I asked the question, where are all the planes? And I took a slightly different tact to Jason. What I was most interested in was the actual path that flights were taking as they passed overhead, so rather than the plan, the actual positions they were in. And it turns out, modern airliners broadcast that stuff from their nose cone periodically over the air in the clear, so that stuff's floating about around us all the time. If you have a little USB digital TV dongle and some free software, you can track all the planes that are overhead right now and capture that information, which is epic in its own right. I think that's wonderful, that data's just out there and it's floating around. I didn't do that. I wasn't sitting up on these hills with an antenna for two weeks, right? That would've been fun, but no. I settled on the same service that Jason initially looked at, which is Flight Aware, which basically allows me to time, place and shift that flight tracking information, so they give you access to historical flight tracks. And here's what that looks like. So that spaghetti is about 400 tracks. It's about two weeks worth of traffic into Glasgow Airport, which is the left-hand pin there. It's a little inaccurate. You can see there are flight tracks that look like the end in the middle of the river. They didn't actually end there. They're just not very accurately portrayed and in fact the data's not perfect. However, that got me the first part of the question. I now know where these flights are and you can see that my little preferred pin up at the top there is pretty heavily overflown, so it was a valid concern to begin with, right? My next question was, okay, so how loud is a plane? What is loudness? So at this point I was kind of concerned that maths and physics limitations might be a problem, but it turns out like the sound pressure page on Wikipedia is superb, is extremely good. It's also got a list of sounds and their volumes which was quite important to me. I learned a little bit about the concept of sound pressure level. I'm not gonna go into the detail of the mass of that, but intuitively, here's a number line. You can see day-to-day sounds off on the left and then a jet engine noise way off on the right around about 120, 130 decibels. And some wonderful soul has put very calm room into that table on the Wikipedia page. That's a lovely thing. I love the internet for that. There is another source of this information that says that the European Aviation Safety Agency actually records the volume of planes as they go overhead for various types and publishes a database. Their number is slightly lower. It's like 90 to 100 decibels sort of in the middle there. And that's the number I chose because it seemed a little more realistic and potentially more scientific. So now I know where the planes are and I know how loud they are at some distance. You'll notice that each of the values here have a number, a measurement, and a distance because with distance, the experience of sound changes. It dissipates the further away you get from the source. So I needed some way to work out how that worked, what the model was for that. At this point, the concept of acoustics and the maths that go with it sort of made me a little bit uncomfortable. But luckily, there's also a wonderful, an extremely good section of Wikibooks on outdoor sound propagation which has exactly this information in it. So I put the maths degree aside for a while and I can do that another time. And I found two important parts of that discourse which I used as my sound model. And they're about losses. So with distance from a sound source, a sound will dissipate. There will be losses associated with it and the two types are spreading losses and absorption losses. There are other losses, but these are the two core ones that I paid attention to. The spreading losses is basically each sound at its source has a given intensity and as it spreads over space, it expands and that energy expands over a shell and the shell expands. So each point on that shell has a smaller and smaller part of the original energy. The larger that gets out. So the folk at the back probably can't hear me at all without the PA. The folk at the front might hear my voice because of the amount of energy that's dissipated throughout this room when I speak. So that's a geometric spreading loss, fairly simple to calculate, which is lucky. And then the second half is absorption loss. So when you produce a sound at source, it travels through atmosphere and as it wobbles the atmosphere, the atmosphere wobbles back, it pushes back and actually absorbs some of the energy for that. And that requires you to consider the frequency of the sound and some atmospheric conditions and so I made assumptions in that direction which basically meant I made a guess at Scottish weather, which if you've ever been there, you'll probably already know about. And the end result was a very simple sound model. And so if you put all that together, I've got where the flights are over a two week period-ish, I have the ability to calculate a so-called distance field. So when you're generating a map of this stuff, each cell on your map represents the distance to the closest plane overhead. So I'm standing on the ground at position and the closest plane overhead might be 10,000 meters away. From that distance field, I can use this little sound model to work out how the dissipation of sound works to give me a view on what the volume of experience sound overhead is. And then, well, at this point, presumably it must be time for a heat map, right? That's gotta be the thing you do next. And so I did, here it is. So finally, I get to answer my question. Right up at that top pin there, you can see the blue and green color and the fact the blue, green and yellow colors are down below that volume of a car at 10 meters. So that's way below anything I would consider an issue if I was going to live there. So question answered, right? We can absolutely move to not live campsy in the campsy fields, which is awesome. Now, there's some inaccuracy here and you can see that those red regions which are where I might be considering not living because it's touching the volume of a plane floating overhead at 5,000 meters. Those inaccuracies are probably due to, at some point, storing latitude and longitude in 32-bit floating point numbers versus 64, which is about five to 10K worth of error. 10 kilometers of error is quite a lot when you're building a map like this. So anyway, mischief managed. Question answered. Thank you, organizers, and thank you for listening.