67: Safety-Critical Software | System Reliability for Live Events with Lars Olson

Brittany Ellich (00:00) Welcome to the Overcommitted Podcast, your weekly dose of real engineering conversations. I’m your host this week, Brittany, and I am joined by

Erika (00:08) Hey I’m Erika

Brittany Ellich (00:09) We met while working on a team at GitHub and quickly realized we were all obsessed with getting better at what we do. So we decided to start this podcast to share what we’ve learned. We’ll be talking about everything from leveling up your technical skills to navigating your professional development, all with the goal of creating a community where engineers can learn and connect. Today on Overcommitted, we are joined by Lars Olson, a full-stack engineer at TAIT the live event technology company behind the staging.

automation and show control systems for some of the biggest tours and spectacles in the world, from the Olympics to major arena tours. His work sits where software meets the physical safety critical side of live entertainment, where code controls machinery moving over and around live performers. First met Lars at CascadiaJS, we’ve met at the last two now, and every time I hear about his role, I get really excited to hear more. So

Hence why we are now on the podcast. So welcome, Lars.

Lars (01:07) Thank you, thank you.

Brittany Ellich (01:08) to kick us off, what is one thing you are currently building or obsessed with learning right now?

Lars (01:13) Yeah, so I I I had a couple things written down. honestly there’s there’s just so much going on at work and outside of work. at work right now we’re doing a whole rearchitecture of our of our software. So it’s about a 10 year piece of software. We’ve decided to redo it at this point. so maybe we can talk about that a bit later. other than that, I have a personal project outside of work called My Hometown Recipes. I think I remember mentioning that to you potentially last year, but it allows you to either use AI to identify recipes from a specific part of the world.

or user submitted recipes, but kind of hard to get people to use a site without people submitting a recipe yet. and then other than that, I am working on a Expo app for my sister because she’s had a nice mobile idea for a couple years. And then I just actually picked up a consulting website for a group of physicists. So quite wide range of side projects at the moment.

Brittany Ellich (01:59) Yeah,

you could say that you’re you’re quite committed, overcommitted even. Yeah, that’s a lot going on. That’s great. Cool. so yeah, tell us a little bit about what you do for work and like how did you end up in this space where you’re writing code for live shows?

Lars (02:02) Yes, overcommit. I am overcommitted at the moment. Yeah. But yeah.

Yeah, so at TAIT, like you mentioned, we’re one of the the largest stage manufacturing producers in the world. we also do like cruise ship entertainment, like theater theaters, opera houses, anything like that. we have two pieces of software, one called Navigator, which elec does it’s our oldest piece of software, it’s about twenty-five years old. that’s generally what most of the the industry uses at the moment. We also have IQ, which is what I work on. that’s been developed over about the last ten years and it’s essentially an easier version of Navigator.

I’ve been with the company now about four and a half years and Navigator’s still hard for me to use sometimes. ⁓ yeah, it’s it’s it’s pretty intense piece of software because it’s just adding features for twenty five straight years. and so IQ was supposed to be this new solution, basically integrates the same with Navigator, the same safety code and everything, but you can train somebody on two weeks instead of two years.

Brittany Ellich (02:51) I don’t know.

Socks.

Lars (03:05) At least that’s the general premise. And so yeah. I gener I developed the front end of that and also generally we call the the f back end of the front end. So all the APIs and everything that integrates back to Navigator. So yeah.

Brittany Ellich (03:18) Okay. So IQ sits in front of Navigator, essentially. Okay.

Lars (03:21) Yes. Yeah, pretty

much just like a a new front end layer, windowing and everything. We have like a existing C plus plus app with a React front end, and then we’re switching to a Rust backend like windowing with a React front end as well. Yeah.

Brittany Ellich (03:32) Nice. Rewrite

it and rest. That’s everybody’s dream.

Lars (03:36) That’s kinda where everybody’s going. Yeah, we had quite a Rust fanboy when I first joined the company who’s who’s won me over over the years.

Erika (03:42) So it’s it’s the like the operators who use the software and it like integrates with the the systems that are put up.

Lars (03:50) Yeah.

so either we’ll manufacture the hardware ourselves, so it’s like the drives that will be an axis going up and down, left and right, return table spinning, or we integrate with pretty much every other solution in the market. And so Navigator will do the safety code integration with the hardware, and then we’ll write the the front end on that to integrate with Navigator and then like you said, it’s operators who will either do a live move or pre program a show to be like at ten seconds move this up and down or

T do time code with certain songs.

Brittany Ellich (04:20) Makes sense. Yeah, that’s really interesting. I think it sounds like something that you probably can’t get wrong super often. It’s probably pretty important to like work well. so yeah, that’s I feel like that can be a little bit different from a lot of software out there where it’s like, it’s got a bug, I’ll go fix it. ⁓ yeah. Yeah, not not quite as much of a thing.

Lars (04:32) Mm-hmm.

Exactly. it the website crashed, just refresh. But it’s like

Brittany Ellich (04:46) So when you tell other engineers what it is you work on, what is it that they usually like assume about it or they typically get wrong about it?

Lars (04:55) The first one I always love hearing is like, I work for stage automation, like whatever. And people are like, I love the lighting you guys did. I love this concert. I’m like, We do everything but lighting. That’s like the first thing people go to, and we’re like, Okay, everything else. ⁓ and that’s like the stage automation that goes up and down. We do performer flying, we do stage like pyrotechnics.

Brittany Ellich (05:03) Yeah.

Mm-hmm.

Lars (05:14) And then a lot of the ones as well, people assume are like, I work on stage automation. They’re like, you write the shows, like I went to Taylor Swift, that’s very cool, you did the programming for the show. And I’m like, no, I work on the application, like the internal tools behind it, and then somebody else at our company will actually do the programming. Yeah.

Brittany Ellich (05:30) Okay. Is it permit

pretty much entirely within your company that they do that, or is it like a a tool that anybody can use across the industry or?

Lars (05:37) Over the last

few years, we’ve actually been kind of slowly building an internal team that does that. But over the last 25, 30 years, when Navigator was first created, it was mostly contractors. And I think it’s hard to break in a navigator in IQ. It’s like generally you you have to be involved with the industry already. ⁓ But for the most part, there are some contractors using it, but it’s it’s generally mostly an internal tool that will send a few people out for each tour to operate it.

Brittany Ellich (05:54) Mm-hmm.

Hmm. Yeah, that’s really interesting. Yeah, I could see how people think the lighting first, that’s like the first thing that comes to mind mind when I’m like, designing a show. But there’s so much I feel like that goes into it that you don’t even realize is happening. like you said, like the stage automation and the stages moving up and down. I feel like when it’s done well, it’s not really a thing you’re like, look at that stage move up and down. It’s just like part of the experience of the show.

Lars (06:10) Yep.

Mm.

Yes, exactly. It’s

it’s completely transformed how I go to shows now. It’s like the first thing I look for. It’s like, look at that really cool thing. And like generally I think sixty, seventy percent of the times it’s actually us doing it. There’s very few shows now I go to where it’s not a TAIT concert. So now I’m always looking for like something going wrong.

Brittany Ellich (06:35) Mm-hmm.

Interesting. ⁓ that’s cool. no bug reports.

Erika (06:45) ‘Cause

I mean I can think of so many things that could go wrong, right? Like it’s like the like the hardware could malfunction or there could be a software bug. Like we’re talking about like designing for failure scenarios. There could be like a really crappy network connection. Like, I don’t know, some of these some of these stadiums or whatever, like maybe they’ve improved their Wi Fi, but I know like I’ve been places where my Wi Fi doesn’t work or something like that.

So like how do you manage that like the network connection and that kind of stuff?

Lars (07:11) Yeah, I we we’ve had

Yeah, so generally we’ll like you said that we’ve had network dropouts that have caused issues before. but generally we’ll set up our own like network within a stadium. And so we’re not reliant on the stadium that we visit. Just generally as well, because every single stadium we go to, every single arena is completely different how they do technology. And so you can’t really rely on the venues you’re going to. but yeah, generally we’ll we’ll set up our own network. everything is generally com completely isolated, so you don’t want like somebody in the crowd trying to hack

into your system mid-show. ⁓ yeah, and so we’re we we we try to be entirely, you know, offline as possible, but a lot of cable runs and everything, so we really don’t even have a much of an exposed Wi-Fi for it.

Brittany Ellich (07:47) Makes sense.

Okay, so there’s a lot more that goes into it. I feel like that is this like are is TAIT usually like one of the first groups that shows up to set up for a show? I would ‘cause I imagine a lot of that stuff you’d have to build around that.

Lars (08:08) Yeah, so w we’re out all the way actually from design of the show now as well. there’s another company we recently acquired that does the what the what the show will even look like, and then they’ll move to the next stage where we manufacture the show. and then it goes all the way to us doing the load in for it and the loadout. and sometimes it’s the entire show, stage automation, like even helping set up the lighting, and then sometimes it’s a combination of like somebody else did the stage, we just did the automation in the background. but we’re generally always like the first ones in there helping it set up.

Brittany Ellich (08:36) Did you was this the first place that you’ve been then where you’ve had like where things have to work like the first time they you know, they that it runs? Is that a new or you or?

Lars (08:44) Yeah. Yes.

That’s pretty much new for me. So like I s after I graduated college, I started out at Navient, which is a student loan company. So sorry if you guys have student loans. I was helping out rewriting the COBOL mainframe system to some a modern stack, which was Kotlin and React. I was actually hired as a back-end developer and then quickly realized that I don’t like backend development.

And transition to be a friend of the developer. And so I I kinda learned React there for a few times. I had my own side projects. and then TAIT actually ended up just hitting me up on LinkedIn. And I was like, Well that’s the coolest job ever. How could I not apply? so that’s pretty much how I ended up there.

Brittany Ellich (09:19) Yeah. Yeah.

That’s awesome. Yeah, very different. S I mean, still things need to be correct, I guess, when you’re working with financial systems, but it’s a

Lars (09:30) It

it is correct, but it’s like not a it’s just like a a piece it’s a website, it it can crash and everything and it’s fine, especially working on the front end and everything, but there’s not some like like this talk has mentioned, there’s nobody performer flying or there’s nobody on a stage in front of me at the moment.

Brittany Ellich (09:45) Yeah, that makes sense. has that changed d does that change the way that you design or test things? Like is that does that show up within the work that you’re doing day to day then? you can’t really like redo a thing. so yeah, how does that work?

Lars (09:58) Yeah, I think

we’re we’re pretty tedious about obviously writing unit tests, integration tests. Like we have a pretty pretty end to end stack of of everything being tested. we have a variety of like code review for multiple developers, then it makes it to like one round of QA, and then it makes it to a next round of like full end-to-end testing for QA. we have physical hardware in the office, so we have like physical things we can move up down. and so it makes like even though we try to work quick, obviously, it’s like

ten rounds of of testing before something actually makes it to a customer.

‘cause it like you’ve mentioned, it can’t go wrong generally, and I’ve seen in the past where it has gone wrong. and we’ve bitten curled pretty hard. So

Brittany Ellich (10:35) Yeah. Yeah.

You were mentioning before we were chatting that TAIT’s been around for a while now, right? It’s forty or fifty years. That’s yeah. I feel like that’s also not something that everybody who works in tech is like experienced with.

Lars (10:49) Yeah, it’s it’s kind of a fun one. So like Navigator I keep mentioning has been around for about twenty, twenty-five years now, but before then we were generally just a hardware company. we we started out just manufacturing stages essentially or different props for a stage or anything. I think we our tagline is that we’ve been on all seven continents and in space. so we we did I think it was a Metallica concert in Antarctica. and then when you two did a concert way back in like twenty ten, maybe earlier than that.

we sent some letters up to the space station to like record a video for them and then they presented it during their their their concert.

Brittany Ellich (11:23) That’s cool.

Lars (11:25) But then yeah, just over the time we’ve tried to integrate more and more and we’re becoming like a lot more of a technology first company. I mean that’s just how the world generally is. And like we’re not just doing staging anymore, we’re doing like like theme park rides and like permanent installations which require mostly software. So yeah.

Brittany Ellich (11:40) okay.

Interesting. Theme park is that specifically like the like the traveling theme park rides or or does it matter? Is there a difference between them?

Lars (11:48) No, so like ⁓

generally a lot of it is like we’ll do either like big props. so like you’re on a ride and you’ll see a prop going up and down left or right next to you. So like sometimes it’s not the actual sh show control, but it’s the things moving around you, or like we’ll do scenery building. I don’t know if I’m allowed to actually mention the specific rides that we’ve done. But ⁓ yeah. We’ll do a lot of the like automation for like

Brittany Ellich (12:08) Fair. That’s interesting.

Lars (12:13) If you’re actually on the ride, like d controlling a big arm, that actually has people on it. So that has many layers of like UI safety code written underneath it, r like manufacturing the hardware underneath that and everything. So

Brittany Ellich (12:26) Yeah. Yeah, that’s really interesting. and there’s this thing that’s been happening recently in the tech industry. You might have heard of it. AI. it’s been mentioned a few times. how is that like I I know like the c the talk right now is always like, you gotta ship faster and move things faster, but like that doesn’t necessarily work where you are. Like how how is that working where y what are your thoughts there on AI?

Lars (12:49) We’re

trying to find a good bounce for that. I actually personally love you love AI. I was a big AI hater like a year and a half ago and then I actually used it and I was like, okay. but no, our our director for software engineering is like a huge proponent for it. Like we have Claude, we have Copilot and everything at work and we’re allowed to use it. generally it’s still gonna be this incredibly strict like QA process, still requires human developers re reviewing the code. We still have actually copilot reviewing our PRs now.

Brittany Ellich (12:53) Mm-hmm.

Lars (13:14) But generally I try to break it up between like what I’m gonna use AI for is like I generally don’t like to use it for if it’s gonna be a something’s as like a safety system. if it’s actually interacting and it’s something that can trigger an emergency stop on a device. It’s something I’d like to be a little more handcrafted for. but generally it’s been great. I I actually have a Claude code thing going right now on the side at work. ⁓

Brittany Ellich (13:38) Thanks.

Lars (13:39) It’s yeah, it’s great. So I I love doing it. I mean there’s a lot of stuff that we can work on that’s just like simple displaying of information that is perfect for AI.

Brittany Ellich (13:45) Mm-hmm.

Yeah. Yeah. No, same. I almost always have something running now, it seems like in the background, just operating. it’s a very, very weird world. and it sounds like I mean you have a lot of controls already built in for safety anyway. So really that might be one of the most ideal areas to be experimenting with AI coding, I would think, because like you have a lot of stops in place, it sounds like to

Lars (13:54) Mm-hmm.

Brittany Ellich (14:11) review and make sure everything is good.

Lars (14:14) Yeah, and and thankfully for like the hardware, we have like a physical piece of hardware that goes up and down. But then on top of that you have like safety PLC code and controllers that does its own level of safety. Then you have Navigator doing safety code, and then you have us doing our own safety code for the UI. ⁓ and it really is like that required full stack because you’re actually moving people, you’re doing things in front of a crowd. Like there’s there’s certain cases though where I’ve seen like really early on, when I joined the company about five years ago, we had some persistent web bug that kept

Brittany Ellich (14:28) Mm-hmm.

Lars (14:43) crashing a window that would actually show you where you’re like, let’s say you’re in the middle of a show and it says, I’m at position 10, I’m going to position 20 next 20 feet. the window would crash every time you’d start up the software. Even if we had like a specific error boundary, you’d refresh, it would crash again. So it’s like that’s just not a well-designed piece of software because it’s like, how do you get out of that? The only way is booting up our old piece of software navigator that you may not know how to use if you’re trained on two weeks of of IQ.

Or you can press the E stop button and stop the concert because it’s safe, but then your show can’t proceed. So.

Brittany Ellich (15:14) Yeah.

Yeah, you can’t

Erika (15:16) So how did

how did like how did you respond? Like how what was the what was the learning like from that and like how was it how was that handled?

Lars (15:25) Yeah, so like when I first joined the company it was strictly JavaScript, no TypeScript on the app. so we had a just a lot of interface bugs of just like using things that weren’t null checked, or using incorrect like we used manipulation on a number and it’s actually a string. and a lot of like fixing small bugs just like that, thankfully, was just just conversion to TypeScript, which most most of our app is now. and then a lot of it was just like really bogging down and like we created a much higher QA stack.

So like all this QA I keep talking about wasn’t fully in place five years ago. so redeveloping how we did QA, redeveloping how we did unit testing, integration testing like that was was kinda how we solved it.

Brittany Ellich (16:03) That’s cool. I think that that’s work that at least within my career, maybe it’s just places that I’ve worked. But I feel like it’s always a thing you really want to do, but don’t often get the the time set aside to do it because everybody’s always like, no, new features. So that’s cool that you’ve got actually gotten to see how that is implemented over time.

Lars (16:12) Mm-hmm.

Yeah, this

Yeah, and and we have these these crashes as well that occur and it’s like, okay, it’s it’s crashing and we have another show tomorrow. Like this is an opera house, for example. So it’s like the bug needs to be fixed today and we’re gonna ship it tonight and it needs to be used by people tomorrow. So it’s like we need to work slow for safety, but we also have these incredibly tight deadlines. So that really makes us need to be diligent when we’re actually developing it the first time, or we’re gonna run into these scenarios where it’s like, Okay, well, you might need to work a little bit overnight.

Erika (16:22) Yeah.

Yeah. Yeah, that it is informative though, like of this idea of a lot of small bugs cropping up and creating these larger things. so that’s a good lesson to learn. Sometimes it’s not one big bang, it’s a lot of little things that add up.

Lars (17:06) Mm-hmm.

Brittany Ellich (17:06) Yeah. I feel like you probably see a lot of really weird failures too that like somebody who’s just works specifically in web development wouldn’t actually anticipate. like, somebody, you know, spilled beer or something on like the the stage or something like that. Is that anything weird that you also have to be prepared for that you wouldn’t need to otherwise?

Lars (17:21) Exactly.

Yeah, this is one I’ve actually run into in the past, is let’s say a piece of hardware goes bad. I mean, it’s just like a drive suddenly is shorts out or something. It can still act like it’s going correct. It can perfectly spit out data, but it’s just gonna start spitting out bad numbers. And so it’d be like, let’s move to let’s move to position like feet twenty, and it’s like, okay, I’ll go to twenty, and it’s like four hundred just because the hardware’s gone wrong. And so, like, how do you code around that? I mean, I don’t even know if I have a great answer about that. It’s like the safety code will prevent the device.

Brittany Ellich (17:46) no.

Lars (17:54) From going past its specific measurement, but you do also have to catch this in the UI being like, this is incorrect, it’s just displaying wrong information. we do get statuses back from devices, so like maybe it’s a lot of like conditional rendering, like maybe we don’t render to the current position if we can see the devices in a bad state. so you’re not confusing the user.

Brittany Ellich (18:16) Yeah. Dealing with non-deterministic outputs before it was cool. That’s funny. cool. Well, I want to talk a little bit about like what people that are working not on safety critical work can like steal from some of your practices because it sounds like there’s a lot that you’ve learned in this role. so one of the things, is there anything that you think like somebody on like a very conventional

Lars (18:20) Mm.

Brittany Ellich (18:39) web development track might be able to like learn about working on live events or working on hardware or anything like that with physical stakes.

Lars (18:48) Yeah, I think the biggest one I’m always going to mention is error error boundaries have become like my best friend. generally you’re just gonna try to at least narrow it down to the smallest piece of your software that can crash. if that’s an individual widget, if that’s an individual certain section of the page. I’ve I’ve dealt with a lot of that w before. Like in the past, if our web crashed, the whole every single web page would crash. then we started isolating it down to a single web page, then a single section of the web page, and we’ve just kind of gotten more granular as time goes on.

Which has been great because I I’ve started taking that into my own personal projects now, like this recipe app that I mentioned. I had a lot of bugs with the specific map section.

I’ve tried to error boundary like certain parts of my own app now where it’s just like, okay, this can crash, but it’s fine, we can still recover the rest of the application. We’re not like giving a bad user experience to other people.

Brittany Ellich (19:32) Interesting. Can you describe what an error boundary is? So you’re basically like preventing an error from going outside of a boundary.

Lars (19:38) Yeah, you’re essentially

just wrapping other components, being like, okay, if it’s gonna catch a crash that anything that occurs underneath this, ⁓ you can r you can provide some type of fallback component to render hey, this this thing crashed, please press refresh to refresh just this individual section. you can even provide like fall like a suspense fallback to show a spinner as it’s trying to reload and everything like that. So

Brittany Ellich (19:45) Mm.

That makes sense. Yeah. I feel like it

Erika (20:01) Yeah.

I was gonna say that that’s the first thing that kind of like came to mind when you were talking before about your work and you know, like the layers of errors, like like I guess this is slightly different than preventing errors from spreading, but with your example before of the

like the arm moving four hundred instead of two hundred feet ‘cause something went wrong. Like there’s multiple places to handle that too where it’s like, well you have you mentioned like a safety check to say like, no, don’t actually move this four hundred feet. Like you’re only allowed to move wherever.

And then also on the like controller side to say like, hey, this is like this this event fired, so like don’t trust this from now on. So you kind of have like two different two different information sources to yeah, to to correctly handle the situation.

Lars (20:56) Yeah, it’s it’s all interesting problems ‘cause like coming from a maybe less like I came from a traditional web sense, but it was also still like internal tools. I’ve never actually like worked on a public website, at least full time. but then going to this company where it’s like it needs to be right the first time as much as you can, really changes my perspective about not just working here, but I think any like any place I work in the future is like how safety critical you need to be, even if it’s just a basic website.

Even if it’s just to provide a good user experience to someone.

Brittany Ellich (21:26) Yeah, yeah, for sure. I feel like that’s something that we thought we talked a lot about, especially at GitHub with so many like moving pieces and parts, like you don’t never want to be the one app that takes down the rest of the site. Always isolate things and make sure like that’s very good, very good.

Lars (21:36) Yes.

Mm. Yeah, for sure.

Brittany Ellich (21:43) are there any habits or any like instincts that you think that you’ve picked up while while doing this work that you’re gonna, you know, for sure take on to future roles?

And it’s okay if not, I mean

Lars (21:54) You I’m not I don’t really actually have anything good on that one. good habits. I think it’s all coming back to like the same stuff I keep saying. It’s just like I I I think what I’ve told what what I’ve talked about on my managers here is like trying to go more the architectural route. And so with AI I’ve done a lot more like spec driven development. And so it’s really like right out an entire spec of of the feature that required. We’re not like asking clo Cloud Co to develop something for us.

But really thinking about every little single piece, we start like diagramming and everything, provide the diagrams to Claude. yeah, we’ve just gotten pretty it’s just a lot more of like minute details now than I was like four and a half years ago on on literally every single piece of this.

Brittany Ellich (22:34) That makes sense. Yeah. And you were saying too, like just getting a a vibe of like how safety critical things actually are. Are there other things now where you know like all right, this is a thing that’s not gonna impact humans or something? We don’t have to have as rigorous of a scope for it. Yeah.

Lars (22:48) Yeah.

Yeah. I always try to like separate it between like visualization of data versus something that can actually interact. so like one window we have on our on our software is called the device window because everything is considered a a device and essentially it just shows every single device in your app. all that’s really used to do is just like show the current position of things, show the status and everything. And so like that’s less less mission critical because all it’s doing is just displaying information to you. It is important.

but then like we have another page that’s incredibly important that’s actually showing what part of the show you’re running, what’s gonna come up next, what like every single piece underneath that. and so we have to be a lot more diligent on those types of pages where you can actually load things on what’s gonna run for a show. This is how you actually we have things called channels. So like you load something to one channel and this is where it’s running the show, but suddenly let’s say that channel crashes, or you can pause it, you can load something to another channel.

Let’s a joystick and you can move it really quickly manually, just because we got into this weird situation where maybe the thing ran into the ground because we moved too far. ⁓ so stuff like that. There’s just like recovery methods in like certain windows where we really need to be careful, and then some are just like, this is a visualization window. We can be a little bit less diligent.

Brittany Ellich (23:57) Mm-hmm.

Yeah, it makes sense. interesting. And then if there was anybody who wanted to build their career around working in, you know, either in live events or you know, anything else that I I guess you’re interacting with a lot of things that your company also manufactures, what do you recommend they like learn or do to get into that?

Lars (24:19) Mm-hmm.

it’s funny, so most of the company is actually like theater kids. I will say fifty or sixty percent of the people at our company I swear are probably theater kids. ⁓ I don’t come from a theater background and so most of the people in manufacturing, designing, we actually have like seamstress seamstress seamstr you can’t say that word. ⁓ yeah. But doing sewing, we have painters, we have like crafters. There’s so many different types of jobs that we have at our headquarters.

Brittany Ellich (24:26) Nice.

Mm-hmm.

Sorry, buddy.

Lars (24:49) We also do like inflatables for music festivals now. and so it’s even for software, it’s just like mostly a passion more than anything. You can always learn what you’re gonna like what you’re gonna do. but I think just like having a passion for the work, because I have a little bit of a music background, really made me like really wanna work here. and including many of the theater kids that work here is like, TAIT’s the place to go for if you wanna do live entertainment.

Brittany Ellich (24:52) Mm-hmm. Hmm.

Mm, yeah. That’s cool.

Lars (25:17) And

like I was hired as essentially a a mid level engineer, but we’ve hired junior level engineers before and interns and everything. And a lot of it is just like be passionate about the work you’re doing. We can teach you how to do an error boundary. We can teach you about safety code.

Brittany Ellich (25:30) Nice. Very cool. Well, with that, at the end of each one of our shows, we always do a fun segment. and this one’s extra fun because you work on something really cool. So we planned the dream show lightning round. so it’s gonna be a rapid fire, no wrong answers, gut reaction to a couple of different prompts. just what would your dream show be to plan, essentially?

I guess it’s not specifically exactly what you’re working on day to day, but it seems like you’re very involved in music and love music. So this is yeah. ⁓ yeah, which one you would want to be a part of, which one you’d want to like prepare for or go to even. You know, whatever whatever feels right in the moment.

Lars (26:02) You said which one I would plan?

Yes, yeah.

fan of electronic music generally. ⁓ and one artist that I’ve always loved is Daftpunk. they’ve yeah, they’ve split up at this point. They don’t do any pro live performing anymore. So like I I really feel like I missed out on that like era of E DM. and that like I’ve always watched live recordings of their shows. So that’d be awesome to to either be a part of or just make it to one.

Brittany Ellich (26:15) Mm-hmm.

yeah.

Mm-hmm.

Yeah, they have incredibly cool shows too. I feel like I th all of their videos are really neat. ⁓ so if you had a a blank check and you don’t have to worry about the laws of physics for one specific event effect at a show, what would you build?

Lars (26:37) Mm-hmm.

Mm, okay. I actually have recently thought about this problem. one thing we always run to in s in like stage production or anything, you always have to manufacture around like cable management. everything has to move around cables. We’ve got cables snagged on things. So like having some type of like wireless power would enable you to build like very cool stages. just being like, we’re gonna move that thing up there. Nothing has to be attached to it. It can just go up and down or left and right or spin around.

without needing to like build in all those crazy extra things to get cables up to it.

Brittany Ellich (27:15) Yeah, I could see that being really problematic.

Erika (27:17) There’s nothing like battery operated that can do that?

Lars (27:20) I mean we can certainly do battery operated, but like the amount of weight on batteries you’d have to like put on this floating thing and like maybe they wouldn’t last three hours for a show ⁓ for the weight of that many batteries.

Erika (27:25) Yeah.

Okay. Interesting.

Yeah. Okay.

Brittany Ellich (27:35) Hmm. Yeah, that’s super interesting. That’s an interesting constraint. what do you think is the most fun engineering problem from the following? A massive stadium tour, an intimate theater residency, a permanent theme park installation, or a one night spectacle?

Lars (27:53) Hmm. Most

think a theme park installation would probably be the most interesting because it’s the one used by the most people over time. And so like a a big s I mean not massive stadium tour. Actually when you think of like Taylor Swift or anything, I think her thing went on for like six months and each s thing was fifty to a hundred thousand people. So it’s probably millions of people. But then let’s say you build something at Disney World or Universal, it’s like that’s millions of people every single year for twenty, thirty years. ⁓

Brittany Ellich (28:19) Mm-hmm. Yeah.

Lars (28:23) And you can get some really interesting problems to solve for that because it’s like we’re not just a one off ride or it’s not just a stage that needs to last one night, but it’s it’s something that needs to be maintained for twenty years.

Brittany Ellich (28:34) Mm-hmm. Yeah. Plus then you can go on the ride and tell all your friends, like, I worked on that. Like that’s I did that. That was cool. which genre do you think makes for the best show overall? Like what do you think ends up like what music genre do you think is the the coolest to plan for?

Lars (28:40) Exactly. Yep.

Personally, I’m gonna say EDM, but actually it’s probably pop music. I I I will think it’s actually more of like the artist than the genre itself, though. ⁓ so like it really depends on how much pride and like how much passion they’re putting into the the design themselves. Some artists are just like, build me a stage, build some lighting, it’s whatever. but then some people are are some artists I think are involved from the very first with helping design out what they want their s their concert to look like. Taylor Swift, I think, was involved from the very beginning.

Brittany Ellich (28:55) Makes sense.

Mm-hmm.

Lars (29:21) I think she had a hundred, maybe two hundred semi trucks touring on her or concert and they would actually build one stage at one city and start building the next stage at the next city. ⁓ yeah, and so it can get pretty intense, what they’re doing.

Brittany Ellich (29:31) Interesting.

all right, and then I guess this will be the the last one here. What do you think is the most underrated part of a live show that most audience members would never notice, but you notice now, every time?

Lars (29:47) Now that I’ve been to a a ton of shows now related to work. So like TAIT will give us free tickets occasionally to a show, which is like the greatest perk alive. ⁓ and so like I’ve been backstage to a lot of events before they happen and then I’ve been able to like see a loadout at the end of the event, but it’s generally the crew actually. So like TAIT will bring in twenty, ten, I mean five to twenty people on a show, but also it’s generally each each city we go to, we work with stage crew who work in that city. And you’ll have two hundred people that put up

Brittany Ellich (29:53) Okay. Yeah.

Lars (30:14) a giant concert in one day, less than a day, and then they tear it down in like two hours. And so it’s like you never would know if you walk in in the morning and you walk in at night, there was never a concert there to begin with. And yet it’s like this huge live entertainment that everybody loves and then it’s gone to the next city.

Brittany Ellich (30:28) Mm-hmm.

Gotcha. Yeah, that seems like a very u useful useful thing. And yeah, it’s something that you would probably never notice, but that’s the point, I guess.

Lars (30:37) Yeah, like you never

yeah, you never see the crew. Like they’re just hiding behind. They’re usually having dinner while the show’s going on or they go out and go sleep back at the hotel room. ‘Cause most of these time these loadout go from like the show ends at say midnight, loadouts by about three AM and then they’re hopping on the bus off to the next city. Yeah.

Brittany Ellich (30:45) Yeah.

Oof.

Yeah, that’s a that’s a brutal schedule.

cool. well this has been great. Is there anything else that you wanted to say or I don’t know, put out there about it?

Lars (31:04) No, I think generally

you covered it. I it’s it’s been awesome to speak about this. I’ve always wanted to get kind of out there more with like public speaking and talking about more about my work. Like every time I’ve I talk to you or more people at the conference, they’re like, Well, that’s the coolest thing I’ve ever heard of, especially like React and Live Entertainment. ⁓ and so like trying to speak more at conferences, more podcasts and everything, which is great. So yeah, thanks for having me on.

Brittany Ellich (31:18) Yeah.

Yeah.

Yeah. This was great. I’m so glad you were able to make it. yeah, and it was great to hear more about about what you do. where can folks find you online if they want to track down Lars?

Lars (31:34) Yes, so Lars is actually my middle name. So you can find me find me at Erik Lars Olson on pretty much every social media. it’s E R I K. and so LinkedIn is Erik Lars Olson, Instagram, GitHub. I think bluesky is ericlars.bluesky. ⁓ but generally yeah, I think that’s pretty much. And I’ve tried to grab that handle for for every social media except for I think somebody beat me out in Gmail. Yeah.

Brittany Ellich (31:51) Nice.

darn.

Cool. Well, yeah, we’ll include links to all of the all of those as well in the show notes. Cool. Well, thank you so much for tuning in to Overcommitted. If you like what you hear, please do follow, subscribe, or do whatever it is you like to do on the podcast app of your choice. Check us out on Bluesky and share with your friends. Until next week.

Lars (32:06) Perfect.