Aug 22, 2022
Sean Fuin is a formally trained mechanical engineer who specializes in automation with python and dynamo. On episode 59 of Fire Code Tech we discuss Sean's career and roles in the architecture, engineering, and construction industry. Sean gives great tips and resources for professionals interested in getting into coding and automation.
Expert Systems: Principles and Programming:
[00:00:00] Hello, all welcome to the show. I'm Gus Gagliardi, and this is fire code tech on fire code tech. We interview fire protection professionals from all different careers and backgrounds in order to provide insight and a resource for those in the field. My goal is to help you become a more informed fire protection.
Professional fire code tech has interviews with engineers and researchers, fire marshals, and insurance professionals, and highlights topics like codes and standards, engineering systems, professional development, and trending topics in the industry. So if you're someone who wants to know more about fire protection or the fascinating stories of those who are in the field, you're in the right place.
Construction industry. In this episode, we're talking again about automation. This episode specifically talks about automation that can benefit engineers, contractors, but the topic of automation and how machine learning AI and other big technology concepts can apply to professional life in general is really ubiquitous in the discussion of how to become more efficient as a professional.
Sean is a trained mechanical engineer and he gives great tips for people who are interested in the cutting edge of technology. And if you have a proclivity for coding automation, BIM Revit, or any of the above, you're gonna really enjoy this episode. don't forget to subscribe and follow us on social media.
Also, if you could do us a huge favor, give us a five star review on apple podcasts. Let's get into the show. Really enjoy this episode. Don't forget to subscribe and follow us on social media. Also, if you could do us a huge favor, give us a five star review on apple podcasts. Let's get into the. Well, Sean, welcome to fire code tech.
Thanks for speaking with me. Hey, thanks for me. It was good to talk about automation. yeah. Automation. I'm gonna talk about automation. Who's not talking about automation. I don't know. It seems like AI. And I keep thinking about AI and machine learning and automation and how that's going to apply to our industry.
But. Yeah, I don't know. We could just get right into it. If you have thoughts on that. I like have been trying to pick up a little bit of Python in my spare time and thinking about machine learning and it's not so like clear to me, you know, automation part's clear to me how that would apply to our industry.
But do you have any thoughts on like AI and machine learning? What we'll see in the future for those kind of buzzwords for what you and I do on a daily basis. Yeah. I don't know. I've never, I haven't jumped on the machine learning bandwagon. I think it's more of a buzzword. No one really understands the amount of data that you need to make it work.
So we have to get a lot better at other things before we can start machine learning our way to success. We have to right. Yeah. I was thinking like, if you had thousands of projects and thousands of device layouts, you know, like really you need like, you know, like 10,000 plus would even, I bet would be a small data set for some of these machine learning projects.
Right. So you would need so many projects in order for it to be meaningful. So yeah, it seems like just the baseline automation stuff. Yeah. Makes a lot more sense. I got in debate with a friend of mine, who's working at a firm that's really aggressive, progressive, and they are working on starting to implement some machine learning stuff.
And they're just like trying to identify like plumbing fixtures. So we have a fun conversation about that, but I'm like, it's only giving you like this five more percent. And like, so all this work has to go in just to get that extra, like 5% success, but it's like, wouldn't it just be easier if the architect called to sync a.
Standardized way. I think that's the approach that we should take. Yeah, but obviously that's had struggles too, so yeah. That's yeah, to my point, I don't think we can get to machine learning if we can't even call a sync, a sync in a consistent manner. That's true. Yeah. I think that it, as with most things in automation, the true value of it is to be found in like the.
Repetitive tasks that you do every day. And then eventually you can get to these kind of like highminded ideas about really complicated stuff. But I will say, I think there is a lot of potential there. So should maybe like auto routing and stuff like that. Right. You can almost turn that into a simpler, like 2d problem. [00:05:00] So yeah, there's so many algorithms out there. Like maybe that I just heard something about. They're trying to map the what is it called? The galaxy like web and they're using some type of algorithm that's based off of some type of mold, I think.
Huh. And so I saw that like YouTube video on that and I was like, oh, maybe that could work for like auto routing. Do you know? Yeah. That makes sense. I mean, so yeah, that like that auto routing is a really interesting topic. Yeah. It's a really hard one, especially fire protection, right? Oh yeah. Fire protection and fire sprinkler.
You guys usually like the last ones to go. Yeah. It's, it's a dynamic. We talk about a lot in inside the firm that I work for is. Fire protection is such a, an odd burden in the sense that, you know, there's so much delegated design that's baseline in our industry. So the engineers by and large will just like shove off the task of actually routing to the contractor so frequently.
So it just causes more consternation in the field and between disciplines and H V a C guys mad. Sprinkler guy got in first and routed wherever he wanted it to. And then just said tough. You know, even though we do our best to try to make sure that people coordinate contractors coordinate and we require coordination drawings and our specs, but it's always a pain.
Yeah. I personally think we kind of design buildings completely wrong. Right. Like we try to lay out. the architecture first, which of course there needs to be like a form, but we sh like we, you know, as mechanical engineer, I'm used to getting a detailed architecture model before I even start running duct work with ceiling Heights already in the model.
And then right. It's always the battle of, I need more room for my ceiling and then not when I do it where right. If you. Kind of have the form of the building, obviously kind of, you know, generically what spaces are, but then you could start to route maybe like the mains, at least using algorithm and even doing plumbing, fire protection, all that together, and then build the architecture's job should be around billion around the infrastructure.
Right? Yeah. And making it look. you know, have them, alternatively, how do we hide this infrastructure? You know what I'm saying? Yeah. It seems like a chicken or the egg problem. Exactly what needs to come, what needs to come first and speaking with the biology, you know, metaphor. I mean, again, I don't think if you watch a baby grow, right.
It starts with like the spinal cord, like the brain, the spinal cord, and like the limbs and everything grow out from there. You know, so you almost have to like, think of the building as having like the spinal cord for the infrastructure and then build around that. That's an interesting thought. I've never, I mean, that makes sense from a that I, I would think that an engineer would think that though.
Yeah, it did the battle between the architects and engineers over, you know, it, I worked at a startup called I. and it was kind of a disaster and this was one of the reasons why. And so I came in late in the game, but also I was in charge of like all the mechanical stuff and they had me design the mechanical stuff around their structural stuff and around just, it was a modular.
So just around all these co. I was assigned to figure out the mechanical and I kept on saying, right, like, no, the mechanical is you should be designing structural architecture around the mechanical, not the opposite way around. And yeah, the owner of the company never really got that. I could see that I've never been able to.
I mean, at least you could have that conversation. I feel like fire protect. Unless it's just like some really specific instance, you know, like with storage sprinklers, sometimes we get into situations where we're pushing around the structural engineers columns and, and Joyce layout. Because if you can't fit these storage sprinklers into this space, you can't build the space.
Like you can't sprinkler it. You can't get a code compliant, build it layout, but arm that. One of the things that modular was like, alright, how are we gonna bring fresh air in? Like, [00:10:00] that was the number one question, at least in my mind was like, how are we gonna bring fresh air into this space? And so my idea was to go to the outside wall and now we, again, we're trying to modular and keep it all tight.
So then it was like, okay, we, so we're like, we're trying to design these things, the modules, right. So like how can we have repeatable pieces that come together? Right. So like a kitchen module might. A closet in it. And my idea was to put, like, try to get really small HVC equipment, but to maybe go in that closet, that would act as like the also act as the washer and dryer closet.
And then we'd have to go out in the wall. Well, my idea was like first put ideally where the H HP equipment should be and then build all the structure around. They were doing the opposite where here's the structure. Tell me where I need to put the H V C. So yeah, those working with those seems like there's always constraints, but the more heavy the building in, or site constraints, just the more difficult the entire process is.
So I think modular is like the extreme of. Constraint. I mean, you're like the whole idea is to get it as confined as possible or, you know, just kind of these succinct little pieces. So that's, it's interesting. It's almost like putting right. If you always have the analogy of like, if you put, if you're trying to like pull of Legos back in a box in ways, if you put all the small Legos first, then the big Legos aren't gonna fit.
Yeah. Okay. So it's almost like that with the structural, the architecture first before the H V C. Hmm. Yeah, I think that's a interesting way to think about it. I guess we'll just have to get faster in the automation and the prototyping to where that is. Like we can have a conceptually generated layout that's quick so that we can raise that red flag fast, but there's so many options for.
Mechanical engineers. I mean like fire protection sprinklers are straightforward. Like you got like what maybe, maybe two most common system types, you know, wet are dry sprinkler systems. There's no like manufacturers really that are driving these big configuration changes. But how much variation do you guys get with building.
I mean a lot, a lot of barriers. That's another huge right thing to tackle. Yeah, for sure. The, the jurisdictional stuff is the biggest variability. Just as far as like military versus commercial versus, you know, you could work for department of defense or you. VA or what, you know, all these different government, high rise Highrise, you know, so all that like has a big impact.
Yeah. Not auditoriums and stuff, but, but yeah. So I know that we we've had for the listeners just tuning in we've, you know, we had some technical difficulties on the, on the first go around. My, my recording program just was given us fits, but I wanted to. Yeah. That's classic problem for technology. I can't open up laptop walls today with rabbit, so oh yeah.
Clap. Wait, what is it? What kind of walls with rabbit? I'm having trouble opening up Autodesk docs models right now. Oh man. Yeah. It's like that one day when everything Autodesk's cloud client went down, it's like, I was like, well, Autodesk never goes down. Surely it's not that. And then you get the email and it.
There whole cloud servers down for a day or whatever. I forget when that was, but seems like there's always something it's harder. It makes things harder. Yeah. But I wanted to let people know a little bit about your background and you could keep it abridged, but, you know, tell, tell the listeners about kind of your working history and kinda how you found your role map.
So I am a mechanical engineer degree. Got my EIT T still wanna get my PE so work on that. Just one more test to take. Yeah. So going through school, I had my first internship at a H V C firm in St. Louis. And I hated it. They were still, they were just starting to play around with Revit. I actually remember, I somehow I got access to Revit and to get the internship, I like pretty much printed out a model of somebody else's work and threw some ducks in there and was like, look, I know Revit guys.
But yeah, I was left doing a whole bunch of CAD stuff there and like, dude, they had me getting photocopy [00:15:00] stuff. So I saw a lot of manual. You know, I was the lowest man on the ton pole. So was a lot of manual tedious stuff and I hated it. Then I got like a year and a half long co-op at a place called tech manufacturing, which they like are a C C shop for aerospace parts.
And so that was really, really cool. So at tuck manufacturing I guess I should back up the. In college. I started learning Excel. That's I think where the story should begin. So in college I started learning Excel and we got really good at it. Me and my buddies that we had to do labs with. So we had labs.
You have to make these Excel spreadsheets and, you know, before, even the lab, we were doing it. So I was starting to get good at Excel. The firm that I had, my internship was used in a lot of Excel and there's this old, old timer. Awesome. Dude taught me even more. He was always showing me little tricks in the.
Excel. So that's been really good, which led me to, I think, in the co-op job, cuz they had an Excel spreadsheet. That was just the most insane thing I've ever seen. So they would bring in so like Boeing or Lockheed would like give these packets of all these parts. And my job was to like find the boing box of the part, right.
For the raw things, get out what material it is. So it might be aluminum might be titanium. The grain, which way the grain runs. You just had to like pull all this information out and then put it into this Excel document. And then at that point it would just kind of do its thing. And you did some other stuff to try to get an estimate on the CNC time, but like do this thing and sew many formulas and so many sheets at the end of it.
It even calculated like shipping cost and all this. And it made what was called, like the red sheet, I think, is what they called it. And it would be like this big review of all the parts and they'd say, yay or nay. So all the higher ups would come in and we'd review all the parts. After all this information's populated in Excel.
And like I said, it's given you cost. It's like it's gave you everything you need to know to say the part is if they're gonna bid on the part, what they're gonna bid and if they'll move forward or not. So I was just like, this is really. So I learned even more Excel there that internship or co-op was up and I needed a job and guess where I landed H VC designer, the thing that I thought I would hate.
So yeah, luckily the time dynamo was starting to become a thing and began my Excel. I was like, oh, DMO can get stuff from Excel into Revit. So that started. And then I realized you could do stuff with geometry and then snowballed from there. Now I've been consulting for three years, four years building custom workflows for clients.
That's very cool. That's neat that you had that saw the power of this
kind of streamlining of, of very difficult and unique process. And so you kind of had that seed planted early in your career for the, the possibilities. Yeah, for sure. Before all that, by the way I worked at restaurants for like my early twenties and I worked under this guy, Vita Elli, who's now the executive chef's chef for Anheuser Bush, but he's like a Gordon Ramsey type.
And he, he was all about, I started with him when I was 16. And so he is all about efficiency. So I. I think that's maybe when it was really embedded in my mind, but then just got reinforced in all these, you know, other little adventures I had. Huh. Interesting. When you're talking about, you know, you make custom kind of like dynamo and Python, like solutions, is there anything like you don't have to get specific?
Specific projects or tools, but like, can you give examples or maybe talk a little bit about your project work and give the people listening some ideas of these applications? Yeah, I mean, I guess there's anytime the big, the first big success I had with dynamo, I think was studying up projects. So I missed a softball game one night and I was really disappointed, but the night was literally spent we behind it was a huge hospital.
I forgot the sheet count, but my, I, I had to work overtime, literally dragging sheets and legends on art dragging views in legends on the sheets. Like it was so silly. So yeah, dynamo again was starting to mature. Yeah, just automating that whole process down to a couple clicks was like, whoa, this is crazy.
So that's first. So any like little [00:20:00] things like that, especially in things that have like a very defined recipe is good for that. So now I've done that with like setting up electrical panel schedules, sheets, and doing a whole bunch of stuff. Like the annotations, like there's like this annotation.
I guess another advanced thing that I did was that was really cool, was working for electrical contractor. So we made a thing called room in a tote. So it'd go into the Revit model and you'd pretty much section the building off by essentially rooms. And the whole idea was we'd get those elements.
We'd make cut sheets for all those elements, fill the materials for all those elements. And then in the warehouse. All the elements, the instructions and the tools, like all the materials, I should say, not elements. So actually the materials, the instructions, how to install the materials, the tools need to install those materials would all go in a tote and be shipped to the construction site.
And that tote would make its way into the room. And then the room would, the tote would open up and the electrical guys would do their installations. Wow. So you were doing like material. Material stock listing basically from yeah. Revit model. Right. So right from directly from the rev model and these guys the company hired a lot of people from the field to do the detailing work inside Revit.
So like they were doing detail work downs, like the bolt. Wow. Which we were pushing our computers so hard to do this. Yeah, it's really cool. That was very advanced. So, yeah, so we'd go through and like make cut sheets, essentially, like get the front view, the side view, the 3d view, the back view, you know, of every electrical panel and then boom, bill of materials.
And then, yeah, there's multiple layers of it. So stuff would build up from like, you know, bolt level to component level. And I forgot that little details of how we did it, but yeah. In other words, you went from Revit model to. All the parts that you needed in a tote that got shipped to the construction site.
That's really cool. That's for people who maybe aren't as aware of like Revit and the common detail level, I mean, like that's like detail level 500 for, I mean, there's kind of like different gradients of how much detail is you can go from like a 2d annotation with no. Information tied to it all the way down to where Shawn is saying where the bolts and the RTS on a piece of equipment are detailed.
And so there's kind of a sliding spectrum and, and that is on the extreme end of like, I'd say most commonly in the architecture, engineering and construction industry. You're at like the middle of that spectrum, probably like 200 or 300. Level of design, you know, you have either 3d models with a limited amount of information or 2d models or at least that's in my experience, what I've seen in the industry.
Well, I like they do have analytical models, right? So I've seen more of that happening where I haven't physically modeling stuff, but you're starting to use Revit more as a design tool rather than a documentation tool. Interesting. Yeah. That's if I add on another thing I've done. So my favorite script I've ever wrote in for mechanical engineering is like zoning a building, and it's just speaking of machine learning so that like uses some clustering algorithms, like canines clustering, and it was a silly problem to solve of just how do you like group these things?
And that was the solution was clustering. Interesting. Yeah. I think that can snowball. To whole bunch of other stuff, but also just allows you to start doing, using Revit as a design tool. Huh? That's really interesting. So you would establish H V a C zones based on your automation. Yeah. So it's all about data, right?
So you have a whole bunch of spaces in a building to zone a building like common spaces usually get grouped together. So spaces like the same thermostat set points, right. Could be grouped together. Exhaust might be grouped together, excuse me. On exhaust fan exterior spaces, right? So north exteriors would be grouped together.
South exteriors could be grouped together. Different windows sizes. Right? If you have a, you know, two rooms next to each other one has a lot more windows, then they wouldn't be grouped together, but right. So you can use all this data to kind of start grouping these things together. And then you're left with like these clusters of spaces that now you have to divide evenly.
And so you can't just like draw a line in the middle. You have to like try to find the right [00:25:00] clusters of spaces. So that's where the clustering comes. And then yeah. So then you can actually do like a decent zoning and yeah, it got to a point where we took this to production and we actually put like sliders on stuff.
So you could limit how much area is on one zone, you know, set a max into that. You could set a maximum to like. The load difference into the spaces you're comparing. There's a whole bunch of sliders on it, so you can start to generate different zone layouts with the same script. Yeah. What I'm trying to get to, what I think is really cool and kind of my north star right now is then taking those zones and going to energy modeling.
So now you could run in theory, right? A whole bunch of different. Run energy models on whole bunch of different systems and, you know, even get you know, upfront cost and life cycle cost and all that sounds really familiar to what I was doing at the manufacturing plant. So, Hmm. That's awesome. It that's such a time consuming process.
I'm sure. Oh, crazy. All those. When I was a designer again, when I was low on the totem pole, just watching all this stuff, I'm like this, cuz they're doing all the calculations to like, you know, decide which system to pick. And then you have to like write, you know, we, we went through system a, B and C and we're picking system a because of X, Y.
and yeah, it was really time consuming. Yeah. So Revit has access now to energy plus, which, which is what I've been working a lot on. That's awesome. Yeah. That's and I can't see, you know, application to this for really so many engineering systems, like, I mean easily for electrical, definitely for fire protection and you know, other disciplines where, you know, Zoning equipment zoning based on whatever parameters you wanna build in, you know, whether it's a voltage drop or, you know, friction loss you know, all that I could see just so much potential for that.
Right. It's kind of cool is too. Yeah. I've thought about it with electrical a lot and I haven't done it yet, but like, you know, taking receptacles and grouping those for panels, right. So, yeah, that's why I seen doing the stuff that I was doing with the electrical contractor. Like when I was doing that, I was really interested in like automating the process before the processes that we were automating there.
Cause we were just automating the documentation and already I was looking at like, how do you guys know what? Like, cause they were having to put in a lot of information, right. To make room in a tote work. So it's like you just keep on knocking. Let's keep on knocking chunks out of the thing. And that's why the zoning algorithm was so exciting to me, cuz it, it was a huge whole bunch of people told me it couldn't be done.
I was like, and I didn't believe that. So, you know, it was just like, it was this huge gap in the workflow and to solve that gap was just really. Isn't that the start of any good story that they told you that it couldn't be done. I 'em there yet. I haven't won the, the automation princess yet, but oh, does that, did they have awards?
Is that a thing? I'm just saying the, you know, you're always, oh yeah. A struggle. Virtual, the virtual automation princess. You haven't gathered her out of the castle. Hopeful it'll happen in my career. Right. I mean, you're off to a good start. That's fun. I like thinking about these big ideas and yeah.
Holistically looking at how can you take the whole process and kind of make it into discreet chunks for everything to run smoothly is a really fun idea. Yeah. I think that's what I'm really good at. I'm made a dyslexic. And so if you read stuff about like dyslexic and what they're good at. Exactly. What we just said is exactly what like typically dyslexics are good at.
And so, yeah, that's kinda like my superpower now is being able to zoom in and out of like these big problems and just make the connections. Yeah, it's fun. I love it. It's solving, it's solving puzzles and problems. All the. Yeah. So I'd like to, you know, hearing your big thoughts on like the future and what you'd like to, to do with your consultancy, but yeah, like what what, what else is getting you excited?
And do you want to do with, with automation or like what kind of gets you charged up on a like a project level or just like an, an automation conceptually I'd like to keep chasing that thread. Yeah, I think I, I mean the whole reason I started the consultancy was not to get rich or anything. The whole reason was I wanted to do what I wanted to do and [00:30:00] work on the stuff that I wanted to work on and I couldn't find that avenue.
So yeah, I, cause I didn't mention this earlier, but yeah, I quit my job as a mechanical designer cause I knew I wanted to like pursue automation and I, I tried to find a job around Dallas. to do that. And I couldn't find one. So that's when I started the consultancy. As far as motivation, I think my Northern light has always been this vision.
I've kind of already described
of getting an architect model and then being able to quickly generate a whole bunch of different options for the H VC. I guess branching that off into plumbing and electrical would be awesome. And then I guess the biggest, the like end game would be, if you could do that and start to optimize with all the systems and architecture and structure all in one algorithm.
Which man there's a right. I'll know we get there, I think in theory, but of course it's just like, like computational time is a struggle. Maybe with quantum computing, right? Throw another buzz. There's computers are getting bigger and bigger and faster and faster, so right. That is how we're doing buildings.
The super computers and just algorithm. Yeah, I guess it gets into like kind your thing of like where the industry's going. And I don't know. That's where I'm, that's where I see it going. That's my Northern light. It's also kind of scary. Right. Once if someone beats me to writing that algorithm, am I a job?
But so far it's been kinda proven not to be the case. Right? So. Automation's gonna take her job south park in reference there . Right, but it hasn't, and we are still talking about, you know, we started off talking about the struggles of data and data entry. And so there's like these two forces, there's these
entrepreneurial spirits that are pushing really hard for automat. But then there's just like this force holding it back due to like, I almost wanna say like people problem with just understanding, right. Just understanding, you know, how databases work, understanding computer science. Building code, I think is holding us back.
Right. I think we need to digitize building code if we're ever gonna. Right. An algorithm like that. Cause right. If you, like, we were talking earlier, if you change one thing that changes the other. Yeah. That's an interesting thought. You know, in something that I've thought about a decent amount, because you know, like if you take a look at this company up codes who is a definitely super somebody.
Yeah. So they want to like make the, kind of take a tech angle on the building code. And there's been a lot of litigation over that. Yeah. That is about saying that was huge. The fact that they won, that was huge. Yeah, definitely. So, and I was watching that closely because that kind of, you know, if you don't have free license to work with that data set, then like, what are you gonna do?
How can you, how can you innovate on top of that? It's like even upgrades though. Isn't really there to what we need. Like I played around with up codes a lot. And it's great. Yeah. I know they're doing, trying to do some compliancy checks, audience, compliancy checks, but not the, so I've been thinking a lot about, and I was, I went to bill recently and I was asking everybody it's, it's hard to even formulate the question, but like, You start having all these databases connecting, and we all know the pictures that you see, but it's like, how do these things really connect?
And then it's like, how do you take that stuff to scale? And so the example is, let's say I have an algorithm that is diffuser is already, I'm gonna make it real simple. That's not how it's written, but like diffusers are placed 10 feet apart, right? That's algorithm supply diffusers based tend to feet apart code.
So then you're reading you go, now you're doing a project somewhere else. And this one says a supply diffuser will be placed in the center of the room and the exhaust diffuser will be placed by the door. Right. Human language. How can you build a system at scale where you can just keep on adding those rules in?[00:35:00]
Right. So next, you know what I'm saying? Yeah. Next rule is. Code or whatever, for whatever reason, rule number three is slot diffuser goes on window. You know what I mean? Yeah. The flow control for the such a big process is really difficult. Snowballs on you so quick. And I ran up diffusers cause I was thinking of'em a lot about this with just diffusers recently and yeah, so I've so yeah, bringing back to bill, I asked a whole bunch of people, this question of like, how can we start to do this stuff at scale?
And. Make a system that you can just add more layers on top of really simple, right? The only person to give me an answer or a path to even start to go down was Ian from high park. Who's actually the creator of dynamo. I forgot his last name. Sorry, Ian. . Yeah. And so they're working on this thing called Hy par, which is really cool, but they essentially build things that are functions and the functions wrapped inside the functions is kind of like that design logic is the elements that, you know, use that design logic that get placed.
And so then their kind of workflows to layer function on top of function, which again is kind of getting back to the idea of. Function a will make floors function. B will divide the floors into offices and corridors and whatever right function maybe Q would be right now put in diffusers function. Z might be right now, route ducks.
Yeah, so they have a point and his thing was talking about is he recommended that I look at at the book, I forgot what it's called expert systems. So I've done a little bit look at expert systems. I think that's kind of a route and that's a lot older, the machine learning by the way, expert systems, the book.
Yes. There's a cool book. You wanna know the principles of programming principles are programming. I mean, it's not, or it's not even just program. It's more of like a way to structure data in a way. Who's the author expert systems, principles and programming. Josh C G I a R R a T a N O author link up in the show notes.
But it's also another thing that I've been thinking a lot about is right. We have a lot, like the turnovers happening in the industry so bad and like, so a lot of senior engineers that have all this knowledge in their head. And, and again, that's another just like building code. Like there's also this database of knowledge.
It just like we have building code that's in written word. Right. And that has to be translated into some type of algorithm. There's also the knowledge in all of these engineers heads that has to be translated. Oh, no, sorry. I keep interrupting. I was gonna say, I read a terrible article today. They were posting, they said 25% of the professionals in fire life safety think that their knowledge transfer is being completed from people aging outta the in industry.
So like, there's just this terrible. Loss of industry specific knowledge. And like you were saying, this turnover, this, this, this exiting of this longstanding knowledge base from the industry is a huge problem. So yeah, I guess back to the question of like, where do we go or what's gonna look like in a few years, I think it's an open ended question.
When I was at built last time, I actually left pretty pessimistic cause. I'm fairly new to the industry on like five, six years experience now and right. COVID happens. So we all didn't see each other for a good two years. It's a tight knit community at that conference and we all come back and we're all talking about the same problem about data and standardization.
And how do we implement this in our firm? And I'm just sitting there.
Thinking if I retire and I'm still having these conversations, I'm gonna be very disappointed in my life. like, you know what I mean? And I see other people there that have been there for have way more experience than me. I mean, there has been some success of going from like AutoCAD to Revit, but yeah, I Don.
I hope we're on like the, you know, exponential path right now. It seems pretty flat, but soon we're gonna take off. So I guess that's to wrap up that question two options, right? We either stay [00:40:00] linear pretty much linear flat as we have been with our efficiency in the AC industry, are all this technology that we've been talking about finally gets us to, you know, go off exponentially.
. Yeah, that's interesting. Everything. We need that exponential to happen, right. With everything going on in the world. Yeah. I believe that it's gonna happen. And to, to me, what makes me excited is just like how I can finally see companies or individuals like you even existing, like the fact that they, you and March do exist and you are making good content and automation and like, and then you have companies that are becoming.
You know huge, critical successes, like D roots and stuff. So I mean, look back 20 years ago and we didn't even have companies doing custom software. I mean, like in our, in the AEC industry for as a business, right. I consider myself so lucky. It's like, it's incredible. It's just crazy. I came in at the right time for like, there to be this opportunity.
Yeah, that's awesome. And so that part about it just makes me think that we're just at the launching point of this automation trend and, you know, 20 years, you know, test it, test it. Mm-hmm wow. You need to look at test it. You should, they should be on your podcast too, by the way. Shout out to them, but they just got 20 million in funding or something.
Oh, wow. Yeah. I mean, that's just one example of last start. You brought up up codes co tools kinda working on energy model stuff I'm working on. Actually. Yeah. Another question to ask. Will Autodesk always be the top dog in the game, man? I stop test fit, audit desk, stop test fit from having a booth at this.
Esque university. Oh, wow. That's incredible. I I was talking with mod about who's gonna be the Autodesk killer or if there was going to be one in our lifetime and he, you know, kept boiling it down to like, whoever can really captured the holistic picture of the building and all the engineering disciplines would be the, the one to roll them.
All right. I think H park there again, Hy Park's another one. They have it. What's really cool to me about test fit. What I've always seen is they're pretty much doing like, like they're doing test fits. That's the name where it comes from mm-hmm . So they're like, you know, you give a site, you draw out a polygon and it just starts like creating multi-families and they keep the ratios between like you know, units and parking stalls and stuff like that.
But what I see is this really simple geometry. And back to my original thing about like how routing MEP stuff, how we're doing it wrong with trying to do it. When it's really complex, we have a really complex maze. Well, what test fit's output is, is just like, pretty much like massing. So then I've always wanted to layer MEP on top of that.
Cause I, I know what their output's gonna be. Like, I know what I'm getting. Yeah, and it's simple. So those two things right there, those two things makes energy modeling a lot simpler makes you know, like auto writing algorithms, a lot simpler. So I see someone like that's why I see with test fit is because they started the very beginning stage, but now they can just keep adding.
One of the things is engineer to build custom softwares and engineers. You don't know what you're getting from the. Right. So that's like the biggest struggle that I see with writing tools for, or writing tools for engineers is right. We all know what the, a architects come through and I'm like writing the input is so variable from what the architect can provide.
Right. It's, it's a very frustrating thing. Again, one of those times where I always get told I'm wrong, but I'm always like, or like they don't have control, but I'm like, write it into your bid execution plan. Hey, architect. We expect this brand to carry our fodder wall rating. Is that how you guys do it?
If not. Okay. Let's change it. Or can you please do it that way? Here's why, cause I'll save you so much time on the back and forth. You know what I mean? So there's, there's countless
countless examples of that. When you go from architect to engineering, you know that the data drop the model transfer. Oh, yeah. It's it can be done just like a hundred different, like a thousand different ways. Why is machine learning when you can just write a simple execution plan, find an architect that understands automation and what's needed.
Right. So, yeah, I don't know. I guess the business models have to change. And I'm not an expert on the business models, so yeah, but it's fun to think about, I think that's funny that you're like, let's start with mechanical, cause maybe it's good to be naive and not think about, not understand the [00:45:00] business models.
Cause then you can come at it clean perspective. Yeah. Yeah. I mean, I think that's awesome. It makes me really excited to see things like test fit. And for those who are interested in taking a look it's looks like it's mostly kind of residential and building development projects for. Developers and kind of this blocking and programming sort of early design phase, like even before schematic design, kind of programmatic design phase, and you know, anybody who is aware of the money involved in the development for these type of facilities would quickly realize how much market capture and money there is for this type.
Service. I mean, it's, it's immediately seems like a no brainer from, you know, whatever three minutes looking at it. I keep on bringing up. Ian Ian told me when, well, like one of the first times was met when he was starting high par was, and I use it all the time in so many presentations. It's like, why are we starting off projects with a blank screen?
Right. And test fit's a perfect example. What we could be starting projects off with what is high bar? What do you keep saying? I don't know what that is. A company. It's hard to explain what high is. I think they're struggle explaining what it is. Another person podcast would love to I'll leave. I'm gonna leave it there.
I tried to explain it earlier. The business model kinda. I've always looked at the business model as like, huh. Cause they're trying to bring like a community kinda like DMO sense to it, but it's like, huh, it's weird. that's interesting. Well, maybe I'll do a solo cast or something about it, but interesting stuff.
Well, I like, you got a lot of different perspectives and people so far that I've never even heard of. So I appreciate all that perspective, but so for people who want to get started. Dynamo or for automation, like where would you suggest that they start or resources that they should investigate? I would start with having a problem and trying to solve it in a simple problem.
Not a hard problem. I say this story real quick, or just thing I learned through consulting. I'll like when I first decide to work with a client or not, one of the big red flags that I get is when someone. I want you to auto route MEP for me, like, to me, that's a huge red flag, because if that's just that simple of a statement, they have not taken the time to think about the complexity of that problem, the amount of data that's needed to even start thinking about that problem.
Right. Cause what are you, where, where are your sources and where are your targets? Is that data in the model? What do you wanna avoid? How is that data in the model? Are we gonna do it in 3d 2d? Why not 3d? Because it's really com you know what I mean? Like, just, if you ask for something like that, that complex, then you don't have a grasp on what it takes to build tools and you don't have understanding of the data are the geometry behind.
So again, if you're gonna, if you're trying to get into automation, start with a task, don't have it be auto routing, have it be like getting data from Excel to rev. That's where I started setting up projects putting legends on sheets, something that's relatively linear, you know? And there's a lot of tools out there.
There. There's VBM Excel. There's dynamo. There's Python. Dynamo. If you're working in Revits probably your easiest best shot get hooked. And then yeah, just keep on learning until you can never stop learning. So I guess that's what adds say. Well, I think that's a beautiful point to end on keep learning to never stop learning.
Cause that says something that I'm throttling myself. That's gonna be the guarantee, right? The Northern star is to.
Northern star. I wanna have the, you know, in star wars when they have like the death star plan, that's what I would, that's what I picture. It's what you want for your flow chart for your process and your dependency. I want, I want a hologram in the middle of a desk where I can flip some knobs and just the MEP infrastructure, right?
Just like you say, the generative design stuff just switches and you get. Again, just like the red sheet at manufacturing facility that tells you the critical information. And then you, we all sit around that desk and we debate and we have conversations about, [00:50:00] you know, the right choice and to get there.
Yes. Never stop learning. Hmm. Cause it takes so many hats to pull that off. Probably hats that we haven't thought exist. Yeah, I like that. Well, that's a good note, Sean. Well, where can people find you if they want to reach out to you about custom rev solutions or if they wanna learn more about your company?
So within email website, YouTube, Twitter usually my name, our Sigma AAC. Cool. Well, I'll throw some links to your consultancy down in the show notes. If people want to come find you, but yeah. Thank you for coming on the show, Sean. Thank you. Good conversation. Thanks for listening, everybody. Be sure to share the episode with a friend, if you enjoyed it, don't forget that fire protection and life safety is serious business.
The views and opinions expressed on this podcast are by no means a professional consultation or a codes and standards interpretation. Be sure to contact a licensed professional. If you are getting involved with fire protection and or life safety. Thanks again, and we'll see you next time.