Marc Culver: 0:00

Welcome to the Public Works Nerds Podcast. Hey, welcome to the Public Works Nerds Podcast, a podcast of the nerds, by the nerds, and for the nerds. I am your host, Mark Culver. I am here with the one and only uh Brian Simmons, perhaps taking the title for uh multiple guests appearances. Actually, I've only had you on once before, haven't I? This is time number two. This is only time number two.

Brian Simmons: 0:29

So I'm but it it remains to be said that I am honored because of my reverence of the title of nerd.

Marc Culver: 0:36

Yes. Well, you are a I mean, you're you're you're one of the top nerds uh that I know, public works nerds.

Brian Simmons: 0:44

That's a that's a huge compliment.

Marc Culver: 0:46

It is, it is. Um, and and I mean that in in as the the largest compliment that I can possibly give in in that regard. But uh Brian is with Boltman Mank. Um we met uh while I was working at Boltman Mank. Um we nerded out a lot on the phone uh and in some meetings and that we recorded a podcast episode on drones. Uh you are a big drone guy. I am a big drone guy. Um, but you don't call them drones.

Brian Simmons: 1:19

Uh I mean it's 2025, they are uncrewed aerial systems, but I'm I'm not coming for you. Yeah.

Marc Culver: 1:28

Um, but that was a really good episode, and we're gonna talk about drones today, too. Um peripherally, right? So this episode is is actually on uh AI. We're gonna talk a little bit about AI and public works, and what's we Brian and I had the privilege of presenting a session at the 2025 PWX uh Public Works Expo in Chicago. Chicago's an awesome town, it was a blast Chai Town. It was a blast. Um good wine. Yes, we had some good wine. Um, that's a different podcast, maybe. We we might need a wine nerd. Is this just the link in the bio for that? Perhaps just kidding. Just perhaps. Um but uh yeah, so we did a session, we presented a session together on AI and public works. Uh, we kind of didn't really know. Brian came up and said, hey, we should do a session together, that'd be a lot of fun. Yeah, it would be a lot of fun. What are we gonna do? I don't know. AI. Let's do one on AI. Do I get to tell that story? Yeah, tell the story because I have a much better memory than I do of it.

Brian Simmons: 2:34

I mean, my my recollection is that I came up with the idea for a session, yeah. And and that was one way that we were discussing about highly engaging with such a vibrant national conference at that PWX is, and then I emailed you the the idea for the conference abstract and you ghosted me. So then I submitted it anyway. Probably you I submitted it anyway, and I I gave I gave you an out. I gave you an out. I said, I said, Mark, you don't have to do this with me, but I think this would be fun. And then we dropped it because it was it was out of sight, out of mind, until we both got the email saying, Congratulations, your abstract has been accepted. Uh, I'll let you recount what your next email said to me.

Marc Culver: 3:12

But yeah, I I I mean I don't remember the the specifics of it was basically like oh crap.

Brian Simmons: 3:17

They were expertise.

Marc Culver: 3:18

Uh we should probably like talk. We've like we're doing this, I guess. Yeah, did it. And we did. We did. And you you did you did most of the work, and you as as always, you put together a great presentation. Um, and I just wrote your coattails. Um during the presentation.

Brian Simmons: 3:35

Presentation is what 90% live. Like I just put together the slides, but we hammered out the concepts together. So you you deserve a fair amount of the credit.

Marc Culver: 3:43

Yeah. But what did we do? What was the cool thing that we did to showcase AI during our presentation on AI?

Brian Simmons: 3:51

So I feel like this is sort of the definition of like things that go meta, right? Uh as in we presented on AI in public works, and then we doubled down and we used AI to prepare some things in the presentation. Uh, and and I remember there were a handful of people that scoffed when I got to that slide in the presentation when I said, by the way, this presentation was prepared using almost solely AI. Uh, and and I realized after the fact that I maybe needed to explain myself because I I think there were those in the audience that heard that and thought, well, these guys just phoned it in, you know, and and what in reality happened was I wrote the presentation and then I had AI reorder the slides, prepare a summary. Uh, we used a number of AI-generated graphic uh graphics, yeah, yeah, infographics, images, some cartoony things. Um, and so it was basically all populated using a combination of um Google Gemini and Microsoft Copilot, if anybody cares. Yeah. But then what else did we do with for the cherry on top of all that? So uh after the presentation, uh, we took our slide deck and and only our slide deck, and we uploaded it to again a combination of uh Google's notebook LLM. I think it's now called Notebook LM and Copilot, and we had it generate a podcast script, and then we had it generate the actual podcast. Uh, and and this is the point at which uh I had fully intended to have it generate a podcast using our voices. Uh, and I believe I texted you a link to the podcast with some more expletives that said, holy cow, yeah, I'm honestly a little creeped out by this right now. I don't even know that I want to follow through on having it in our voices. And your text back to me was something to the effect of like, what are we even doing here? Right. We they they don't need us anymore. No, but but so so uh you're leading into we're gonna we're gonna experience this together, right?

Marc Culver: 5:46

Yeah, so you know, we had this great idea, and we actually talked about this uh during the presentation that we were gonna create a podcast. We were gonna we're gonna do a podcast of the session. Uh so we'll talk about a few things that we talked about during the session, but we also teased that you know we were recording the audio and that we were gonna use that to kind of generate an AI uh created um podcast, you know, some voices and things like that.

Brian Simmons: 6:10

So what we're gonna do now is we're gonna play uh a few clips from that 16-minute podcast 45-minute presentation, hours of tears, sweat, and blood, and then AI distills it into 16 minutes of audio.

Marc Culver: 6:27

Boil down, you know, all of our all of our blood, sweat, and tears just evaporated out uh into 16 minutes.

Brian Simmons: 6:36

The I the AI didn't even break a sweat. They didn't. Uh but so this is this is uh just as a warning, this is unmanipulated. Right. This is in its raw form directly out of the AI with a a number of prompts. Um, I played with this a couple of times, but there was almost no delay between the prompts and the generation of the podcast. So there were some things in my prompts, you know, we talk about prompt engineering as a function of AI, uh, giving it instructions on respecting the technical portions of the podcast and also making that I I think I told it to make them accessible. So it does a somewhat good job in some places about explaining things. Also, I'm telling you it's unmanipulated because it says some things that are wrong.

Marc Culver: 7:18

Ooh, it does. I didn't and I didn't listen to the full 16 minutes, admittedly, or I don't remember all of the content. So what what when we can point out what it says that are wrong?

Brian Simmons: 7:27

We'll we'll pause it there because I it'll be pretty obvious. Uh there's at least one where actually the podcast sort of tries to uh coin a phrase and it and it incorrectly labels the phrase. Yeah. So it's it's sort of interesting.

Marc Culver: 7:41

All right. So uh with that, go ahead and uh roll the tape.

Brian Simmons: 7:45

Roll the tape. Let's let's hear what it sounds like. Do you wanna, while I'm pulling this up, are you gonna say anything about the fact that we are currently recording live?

Marc Culver: 7:52

Yeah, so right now Brian and I are actually sitting uh at Mystic Lake Casino. We're not here for the casino, um, we're in the event center uh of the of the casino in Prior Lake, Minnesota, uh at the Minnesota Chapter, uh the APWA Minnesota Chapter Fall Conference. And I don't know how many people are here, but there have got to be like four or five hundred people plus here. Speaking of engaged in vibrant conferences, yeah. I mean, I was real blown away. I came in late, of course. I'm always late. And uh I walked into the general session uh room, auditorium, whatever, and it was I was just blown away. Besides the fact that I had to park like half a mile away in this parking lot. Um at least it's not snowing yet. Right, right. But uh it was just blown away by the number of people that were there, and then you know, and you know, we're we're engineers, we're we don't like people apparently. You know, we don't like to intermingle. So there's all you know, there's all the wallflowers, you know, around the perimeter of the standard. Yes, the standards. Um and but yeah, a lot of the tables were full and it was a really engaged crowd. Uh listening to a keynote speaker talk about by the time I got here, he was talking about um he was in uh one of the World Trade Center towers. Um, and I don't remember the gentleman's name right now, but I could look it up. The cop the the rusher is literally sitting right next to me. But um, we've got a really good uh session here, a lot of great people. Uh we we've got a cool um evening event this evening with golf, uh golf simulator type. Or it's uh is it a simulator or is it like a top golf thing? It's it's knockoff top golf. Yeah, is what it is. Okay. Anyway, all right. So with that, with with the the our the context of our recording, so we may get interrupted.

Brian Simmons: 9:39

Uh I assume there's crowd noise behind us, but I could be totally wrong.

Marc Culver: 9:41

Yeah, I don't know that the microphone's picking that up, but I kind of almost want somebody to come up here and like be what are you doing? And then I'll introduce them on the podcast. I I actually I sort of love the buzz happening behind us.

Brian Simmons: 9:51

I do too. I do too. I want to do this more often. It gives it the live vibe, at least for me, even if they cannot hear it. I I sense the hustle and bustle happening behind us. So all right, so here we go. Here is a uh the the podcast generated by AI. This is AI in the Sky by Mark Culver and Brian Simmons, performed by AI. I think that's the correct idea. I love that, yeah.

AI Host Male: 10:13

Have you ever stopped to think about those really disruptive technologies?

AI Host Female: 10:17

Mm-hmm.

AI Host Male: 10:18

The ones that don't just improve something a little bit, but totally change the game. Redefine an industry, even.

AI Host Fe: 10:25

Aaron Ross Powell Yeah, the real game changers.

AI Host Male: 10:27

Exactly. Right. So let's unpack this today. We're doing a deep dive into precisely that kind of innovation. Uh we're calling it AI in the sky, practical applications of drones and AI. Sounds good. We're looking at how these, well, powerful tools are opening up completely new possibilities for public works, how they're really shifting how our communities manage all that critical infrastructure. And our insights for this, they come straight from folks living this every day. People like Mark Culver, Public Works Director in Brooklyn Park, Minnesota, and Brian Simmons, a principal engineer at Bolton and Menck.

AI Host Fe: 11:03

Right, the hosts of the Public Works Nerds podcast, they know their stuff.

AI Host Male: 11:06

They really do. So you know you're getting practical on-the-ground experience here. Our mission for you today, pretty clear, I think. You'll walk away with a solid grasp of what AI and drones can actually do.

AI Host Fe: 11:17

Yeah, beyond the hype.

AI Host Male: 11:18

It's exactly how they're impacting operations day-to-day, the sort of guiding principles for using them effectively, and some real-world examples happening right now. It might sound a bit nerdy, but trust us, it's incredibly practical stuff.

AI Host Female: 11:30

Aaron Powell And that idea you picked off with disruptive technology, that's spot on. It's not just like a small improvement. It's something so innovative it fundamentally redefines an industry or you know creates a whole new one. Right. Drones and AI, they absolutely fit that bill. Their impact is uh significant enough that you see them featured prominently on things like the Gartner hype curve. It signals a really profound shift in how we look at infrastructure management. Now, before we jump into the specifics, it's really crucial, I think, to set some, let's call them ground rules for AI. Okay. The core principle, echoing the experts in this field, is pretty simple.

Brian Simmons: 12:09

At the ground rules for AI, that's that's a minute 50 into the podcast. So so our slides there were all intro. Uh, but there was a significant amount of sort of I'll say AI ethos things in there from you and I. Um it it it nailed all of the salient points.

Marc Culver: 12:26

Yeah, I think and and starting just like the way it started, um you know, AI, we're we're we're you know, we're calling it a disruptive technology, and I think most people would hear disruptive and think bad, you know, like oh well disruptions are not good. Um, and in this case, uh it's actually a very good thing uh for us. And it I mean it's a disruptive technology in every industry. It's going to be a disruptive technology in every industry. Um, but in public works, uh, we're gonna be doing some things, we're gonna be uh collecting and analyzing data at a level and at a scope that just wasn't possible previously and at a pace that wasn't possible uh prior to this.

Brian Simmons: 13:09

A AI tools seem to have only highlighted for me the amount of data that is just absolutely flying by our eyes at any given time. Uh and and I think it's interesting that it gave the def it gave its own definition for disruptive that was not in the slides, uh, but we used the term disruptive in the slides, assuming that we were going to, and we did, we verbally explained it while giving the presentation. Um, if we need an example here, like disruptive is is what lift is to the taxi industry, right? Lyft and Uber. Right. So so we're sort of you and I are welcoming disruption, I think. And and we think those kind of technologies, they they change things so fundamentally, and that's sort of what AI is doing here. Right. It's disruptive, right?

Marc Culver: 13:52

Yeah, and uh and if we don't have the slides in front of us right now. No don't that's a fail on my part for preparation, but um you know, one of the examples that I'm actually using right now, and and and I talked about this uh at the PWX conference because we had piloted it and and I knew we were going to implement this at Brooklyn Park, and and over the last week we've actually done this. Um we have um signed on with Violytics, which is a they actually the the the core of their business is PCI rating, uh and you know, using machine vision, uh well using photos and AI. I use AI in quotes here, um, to rate uh pavement segments. And they also do some really cool things with um capturing and identifying sign condition, uh, particularly not retro-reflectivity, but they don't get to that level, but I think they can identify like severely faded signs, they can identify um obscured signs, yeah, um, leaning signs, uh, some other you know issues with signs that need attention. Um, they also identify um like sunken catch basins and manholes uh in the pavement, so they'll identify those.

Brian Simmons: 15:14

Um so they're giving you condition assessment on everything they can see. Yes. As your staff are driving your roads. Yeah. And and the initial impetus for this was pavement condition assessment, right? Yep. But then they came to you and said, look at all these other things our algorithm captures. Would you like this too? Right. And and being the smart fellow that you are, you you said, well, if you're gonna capture it all, I'd like all of it, please.

Marc Culver: 15:36

Yeah. So yeah, and so they have a really nice website, nice interface. Um, they're actually set up as kind of a mini uh low feature uh asset management system. Um, I think they have a lot of counties that use it. Um, and and they're the assets that they're really focused on are the assets that are visible while you're driving, and those are the assets that the counties are mostly interested in. Um and and we we are we have a much more comprehensive asset management system, so I'm not using it for that, um, but I'm gonna import that data into our asset management system. Um, but so far it's been really cool. Like we haven't imported it yet, so I'm just using their website, and we've had it since Thursday of last week. Okay, today is a is Thursday. Uh we've had it since uh last week, Thursday. Uh my guys started driving it, and we're we're intentionally driving, we're taking, we're not like, oh, let's mount it to a sweeper or a garbage truck or something like that. Taking a separate trip to do we are intentionally driving the entire city now um in one fell swoop in order to collect this data. And in four days, basically, um my staff has driven 120 miles. Wow. Uh captured not quite half the city. I think we have like two sixty, two hundred and sixty miles of street. Um and like I was uh sitting here this morning uh at the table listening to one of the keynotes and pulled up the website and I'm looking at the data, and it gives me an average PCI rating for the 120 miles that it's captured. Um I can filter, you know, based on PCI uh rating. So cool. Um, so I've got I think it showed me about 11 miles of streets that were under 50 rating. Okay. That's bad, right? Uh that's bad. Yeah, we don't want that. Um and you know, just just the amount of data and analysis that I could do in 10 minutes through this website in data that I've collected in four days.

Brian Simmons: 17:38

So what's what's the concept con context, excuse me, for your excitement here? Because you before this, did you have nothing?

Marc Culver: 17:46

We had trained uh personnel within our street division that would go out and assess uh we've got I think eight or ten maintenance districts. We have that we've divided our city up into I was just in Paris a few weeks ago and they're like Arondismo. Yeah, they're like our own Arondismo uh in Brooklyn Park.

Brian Simmons: 18:10

You know, I I I could I could hear you searching for that. You're gonna tell me which which side of this same that Brooklyn Park is on.

Marc Culver: 18:18

Um so it's it's kind of funny. I I put that together uh as I was looking at it the other day, but um but yeah, so they will pick a maintenance district and they'll go rate the pavements visually rate them. Um and and that's fine, and that served us okay over the years, but the problem is is we get inconsistencies because we're not rating the whole city at one time, right? And so you may see things differently three years from now than the gentleman retires, or yeah, or you have somebody else that's doing the rating. And so you just have inconsistencies then in in how the pavements are rated. Um, and it makes it really difficult to make long-term decisions and plans based on that data, then and because you don't really trust it. Well, I mean, it's it's it's it's a good starting point, yeah. Um, but but it yeah.

Brian Simmons: 19:12

It needs refreshing.

Marc Culver: 19:13

It it needs refreshing.

Brian Simmons: 19:14

You have crews going out and doing things to the pavements, right? That that is part of their regular maintenance, right? However, they are doing things that are extending the life of the pavement, they are preventing water intrusion, all the things that we know as engineers are extending the life of that pavement. And and if you've got just a static rating that now is aged, uh you you had a need to update those ratings and then use those those porous streets or whatever, you know, maybe talk about your scenario. Yeah. Like how you're gonna tackle that from here on out.

Marc Culver: 19:42

Right, right. And and and yeah, I mean, all of the above. Yeah. I mean, you know, um what's really cool, what I love about this is we have four phones from uh Violytics that have the sophomore programmed on it and all set up with mounting, and we actually mount the phones on the hood of the Vehicle that we're driving. Oh. Not in the cab, but on the hood. And then we have a little remote to like start stop. I might be doing it wrong. Um, and they say you get better data that we you could do it in the the cab, but you actually get better data. You get you you the images are closer to the way less hood. Yeah. Yeah. Way less hood. Makes sense. Yeah. Um but what's what I'm really excited about is like we can drive it as many times as we want. And and and we also generate our own street view data. Oh, wonderful. Okay. So we have that data in the cloud, so we can go and look and and see things. But let's say we drive the whole city this year, November. Next year, um, I'm gonna be doing a budget presentation in August, September time frame. Uh, we have just mill and overlaid 11 miles of street. I can say, guys, go out and drive those 11 miles, get a new rating on them, update our PCI rating.

Brian Simmons: 20:52

And and get new street view pictures. And and get new street view pictures. I fully expect you to be flogging that in your presentations to the decision makers that are giving you money.

Marc Culver: 21:00

Yeah. Yeah. And so I'm I'm like super excited about implementing this the this tool. Um I'm I want to learn more about how it's doing it, how it's capturing the data, and how it's analyzing and things like that. There you this will not surprise you. This is a German company.

Brian Simmons: 21:20

Yes. I mean, if they're gonna make Audis and BMWs, not surprising that they figure this out.

Marc Culver: 21:25

Yeah. Um so anyway, I'm just like really excited. And there are other companies besides Violytics that uh do this. I think Civil, uh I think it's Civil. Civil.ai. Civil.ai Y V L offers a similar service with with uh asset management background as well. X Weather. Um NextWeather. X Weather. X Weather.

Brian Simmons: 21:45

Formerly known as Visala.

Marc Culver: 21:47

Yeah.

Brian Simmons: 21:47

Yes. So this is this is sort of a uh wide open market, but you partnered with Violetics. Right, right.

Marc Culver: 21:55

We we had somebody actually I'll give a shout out to Jen Desrud from SCH. Um, she they've kind of developed a relationship with Violetics, and she reached out and said, Hey, do you want to do a pilot for us? And because we want to test these guys out, see if this is something we want to recommend for our clients. And I said, Yeah, bring them up. And we tested, and I think it was in March, and we're driving around and collecting uh PCI data, and it was awesome. And and it took me a while to actually get it implemented. Well, I had to become the director. Um, so I became the director, and then uh your long-term plan to take over the world.

Brian Simmons: 22:26

It hit it. It starts with becoming public works directory, right?

Marc Culver: 22:30

Um, and uh yeah, it I mean it took a little bit of time to get going. I would have liked to have done this in October or something instead of November. They're driving around the because there's some risk there weather-wise, but it's and potential for leaves covering up distracting exactly, right? Exactly, exactly. At least here in Minnesota. Yep. But we have swept our city. I think we're on our fourth time right now. Oh, it's a city. Okay. Yeah. So that's helped. Um, but hey, so getting back to the topic, like this is just like a really great example of how one AI application uh is transforming my department uh at Brooklyn Park. I mean, it's one division, but it's really transforming how we um assess the condition of our our most important asset, our pavement, and um and how and what we do with that data and how quickly we can use that data and make updated, um relevant uh decisions uh on that data now.

Brian Simmons: 23:32

And so data-driven decision-making term just keeps it's it's like circling your head right now as you talk about it. Yep.

Marc Culver: 23:38

But talk a little bit. Talk a little bit more about AI in general. Sure. And uh you had a phrase in the presentation, um, and I can't remember it now, but there was uh, and this might be a part of the present podcast.

Brian Simmons: 23:53

Or is this is this a good spot for us to let the podcast play it and get it wrong? Yeah. Oh, is that what happened? That is what happened. All right, yes, because uh okay, as long as we're leading into this, then I said that uh drones specifically and drones using AI were good at doing things that were uh dull, dirty, dangerous, and repetitive. There we go. And uh listen for the reference in the podcast when it refers to the four D's. The four D. The four D's. I didn't know I didn't know repetitive started with a D, but this is where this is where, like I said, raw and unmanipulated. The the podcast just it rolls right through it. That's funny. Yeah. Okay, so then we'll roll the tape again.

Marc Culver: 24:31

Yeah. So we're gonna queue up that segment.

AI Host Female: 24:35

It's pretty simple.

Marc Culver: 24:35

Yeah.

AI Host Female: 24:36

It must do good.

Marc Culver: 24:37

Do good.

AI Host Female: 24:38

I like that. Yeah. These tools they have to enhance human ability, help us do more with less. They're not meant to replace people. The human element is still invaluable. And when we say AI here, we mean a whole range of things: machine learning, LLMs, even uh convoluted neural meds.

AI Host Male: 24:54

CNNs, right, for image recognition.

AI Host Female: 24:56

Exactly. Those are basically advanced systems built to spot complex patterns in images, perfect for finding tiny defects on a bridge or a pipe, that kind of thing. So it really boils down to responsible use. The goal isn't AI taking over, it's AI serving us, boosting our productivity. There's this great quote, captures it perfectly. I don't want AI to create art while I fold laundry. I want AI to perform those mundane tasks so that I have more time, more space to exercise my own creativity.

Brian Simmons: 25:25

Was that the quote you were looking for? Yeah, it was. That was the quote. I totally mischaracterized you. Yeah. It the other one is coming.

Marc Culver: 25:32

Yeah. And that one still kind of messed up the quote. It's like, just say the quote.

Brian Simmons: 25:36

Don't rephrase the quote. It did it did. It it thought it could make a more efficient use of words. Right. Of my words. Right. And I and in in the slides, that phrase was actually in quotes, and it still changed it up a little.

Marc Culver: 25:49

Yeah, that's interesting.

Brian Simmons: 25:50

It it is interesting.

Marc Culver: 25:51

I mean, it's still scary. Uh let's let's pause here for a moment and just uh acknowledge and appreciate those voices. All right, these are AI generated voices. Those people do not exist. Right. These are not real people. This is not uh your morning, it sounded like a morning show type of thing. It does like a radio, morning radio show type of uh hosts. Their voices are pleasant. They are. They're they're they're not bad to listen to, they're not as good as our voices. Definitely not. Um we're biased. Yeah, and who's who's the guy and who's the girl in in this, by the way?

Brian Simmons: 26:23

I don't I don't you know, I had already wondered that out loud and decided I wasn't gonna bring it up. Let's let's not broach that topic. It obvious it obviously came up with its own back and forth. Yes. So they they have actually more interplay between the two hosts than we did giving the presentation, right? Which I find interesting, right?

Marc Culver: 26:40

Right, which which gives it that it's almost too much. It is almost too much. In my opinion, that's what gives it the kind of that morning show uh radio program vibe to it, in my opinion. Yeah, yeah. Um you live. But but yeah, uh, so that's just okay, again, just scary AI generation. Um, there we we are not necessarily somebody can just create a PowerPoint slide, dump it into this uh application, and create a podcast episode.

Brian Simmons: 27:11

Yeah, so all of the the inflection and the interplay and all of those things is is what the AI brought to it, right? From pure text of our slides and the images in the slides. Yep, because it's it's obviously recognizing some of the things that are in the images. Uh when it made the reference earlier to the Gartner hype curve, I inserted the Gartner hype curve. I made one reference to it. We talked about it, that was related to the disruptive things, but it it it sort of described where things are at on the Gartner hype curve, even though that was a graphic in the slide.

Marc Culver: 27:42

Yeah.

Brian Simmons: 27:43

So it's it's picking up on all of its context clues. Yeah. And then some.

Marc Culver: 27:47

Pretty surprising. But let's get to the quote now. Because uh, you know, talking about AI and you know, uh, one of the fears of AI is uh, you know, that it's gonna take over our jobs. Um and and you know what, let's be honest, let's be perfectly honest. There are jobs that will not be necessary someday because of AI. Right. There are a lot of data analyst positions and some things that just won't be needed, right? Because AI can do it faster and cheaper and better.

Brian Simmons: 28:18

There's probably a whole philosophical discussion here about um human industry that's been created by human industry. Yeah. Right. Like, you know, someone I don't want to continue hounding the idea of data analysts, but someone who sorts numbers into a spreadsheet, right? If an AI can do that, uh, that's a position that we have created to enable something else. And now that person can better use their time, talents, and creativity to do something better, perhaps. But um, that that sounds really judgy. I don't mean for it to, but um the repetition for sure. Yeah.

Marc Culver: 28:51

And and I think to your point, uh and say the now say the quote the right way.

Brian Simmons: 28:55

Um so I I I I don't remember exactly the words, but I I I'm sure I'll get it better than the AI did. Uh, I I want AI to do the mundane things for me, like fold my laundry, so that I can better spend my time um exercising my creativity. Right. You know, so I don't want I oh it started with I don't want AI to create art for me. Exactly. I want AI to perform the mundane things for me. So I can create art. Yes, so I have time to create art. Yes.

Marc Culver: 29:24

And so now it's it's you know, and and going back to this violytics uh example and this pavement rating thing, like my guys were out there spending time, you know, and it's gonna take them longer to drive the street at a slower speed to rate the pavement than it is because they can drive 30 to 40 miles per hour, and they better be driving 30 miles per hour through the residential streets, by the way. But they can drive 30 to 40 miles per hour um and capture this data and have it analyze it much quicker and much more um consistently than than they can do as a human. But now that frees them up to like look at the data and and use their brains to um, and obviously the engineer and myself we're we're also gonna be doing that and making some decisions, but they can make some of those decisions, they can look at some of those things, they can look at that map and say, hey, we got a lot of potholes here, let's go focus crack sealing in this area.

Brian Simmons: 30:21

Or what is happening here, right? What kind of conclusions can we draw from the thing that from the data that we weren't noticing? Right.

Marc Culver: 30:28

And we don't we don't have robots that are filling potholes for us or crack sealing for us and all that would be clear. Well, I mean, we don't have them. Um they they they may be in production, I don't know, but and maybe someday we will. Pat and pending. Do I need to say that? But those guys that were doing the ratings can now, you know, we can actually get more done of actually fixing our roadways for sure. Um so that's just an example of using AI. You know, don't be threatened by AI from a profession perspective. Embrace it and use it to make your job better and more efficient.

Brian Simmons: 31:00

And it's probably worth my saying, too, that those ground rules are are things that we brought to this. Like we we laid those out in the presentation so that people felt comfortable sort of taking these things in, right? That this this was our ethos around the uses of AI. And there are for sure nefarious uses of AI out there, but we're not we're not focused on those. Right. And we're specifically focused on sort of this AI for good topic, which is what they were joking about, and and again, sort of missed the mark a little bit in the humor, but that's okay. Yeah. So let's talk about the the dirty, dull, and the four D's.

Marc Culver: 31:36

The dull, dirty, dangerous, and repetitive. The dull, dirty, dangerous, and repetitive. Uh and you have you provided a couple of examples of that during the presentation, but there's another great application of of AI is to take those tasks that are dull, that are dirty, that are repetitive, um, and and do those for us, so then we can we can do better things.

Brian Simmons: 31:59

Absolutely. So one of the examples that I talked about in this presentation uh is a product that we are partnering with called Sewer AI. Uh their platform is specifically named Pioneer, but I'm I'm not sure that that matters. Uh again in performing condition assessment. Um it's it's something that the visual will indicate the distresses, but it is fairly common for a young engineer to be tasked with uh and and somebody will invariably joke it's your turn to watch dirty movies. Yeah. And when they say that, they mean here here are the sewer division across America has used that joke.

Marc Culver: 32:34

They're obligated to they are. It's it's kind of like you must see say this before you watch this video.

Brian Simmons: 32:40

It it's sort of like how you have to click tongs to make sure that they're working before you use them.

Marc Culver: 32:44

Have you seen those TikTok videos of like a guy versus a girl? Absolutely. Like you lay out the three things, like the tape measure, the tongs, and what was I can't remember what the other item was, but um, you know, and he picks up the cloud, the tongs and he clicks them.

Brian Simmons: 32:56

Yeah, yeah. Yeah. So so it they're they're compelled. They're compelled to. But uh at every young engineer, at least prior to now, has had to watch those videos looking for, you know, distresses and issues. Um I use the term distresses as a sort of a more diplomatic way to say things are could be failing catastrophically. Yep. Right. But um, sewer AI is a platform that has a trained visual model similar to the way you describe the way Violetics' algorithm works, uh, and they will rate those pipes for us. Um, and so we've partnered with Sewer AI to upload a number of communities' data to sewer AI, and uh we get a rating out of it, and then we get links directly to the photos of those, you know, whatever it is. It's a hammer tap, it's a crack, it's a break, it's an offset joint, all the things that plague us in the in the sewer world, right? That allow infiltration and inflow. And so when I say, you know, sometimes not really dangerous, but definitely dull and repetitive. Yeah. You know, reviewing those and looking for those things that enables that young engineer now, instead of spending hours, well, and hopefully it's fewer hours if they have discovered the fast forward button and you know, preferably watch those videos on not one X speed, but right you know, if not, that's where they are. And so distilling that into a report where they can then visit those distresses and you know rank them, or or they can even QAQC whether or not they agree with the automated ranking system. Yeah.

Marc Culver: 34:20

Yeah, and we you know at Brook in Brooklyn Park, we haven't gotten there yet. Um, I'm trying to, as I've talked to my staff um about things that I want to do, and I want to do a lot of things. And if you if you are a regular listener to this podcast, you know that um I embrace technology almost to a fault. You're ambitious in that world. Yes, I love it. Yeah, and and I want to do a lot of things, and there are there are budget constraints uh related to that, there are staff issues related to that. I I my GIS guy, uh Russ, we're we're we just implemented four domains of asset management uh in 2025, and next year I want to do another six, and he's just freaking out that we're gonna do six because he wanted to break it up like three and three. I'm like, no, we're gonna rip the bandit up, we're gonna get this done. Um, is your minion when it comes to change management? Yeah, I take it. Yeah. Well, yeah, I mean he's my my GIS guy, but yeah. Um and but I I recognize that if I'm going to do these things, if I'm going to implement these technologies, that I need to make sure that I provide the resources for my staff. If I can't increase my staff in order to manage these things, and ultimately I'm hoping it makes my staff more efficient. So we have more time to spend on these things. But initially, there's there's you know the adoption of the technology and getting it implemented integrated into our work processes and such. Um, I'm gonna provide the resources to help my staff um deal with these things and implement these things and work through it and deal with the problems that uh and and questions and frustrations that are invariably gonna come up while they start these things. And so we're you know, we're we're bringing on some consultants and uh some outside help um for them to reach out to and get help uh when this happens. But so that's just an example of that. But one of the things that I really, really, really want to do is you know, again, every community has hours and hours and hours of dirty videos. So many. Uh too many sewer videos. And you know, we wanna we all want to try to figure out how do you manage these videos, and IT departments hate us because we've got gigabytes and terabytes of video on the network somewhere, and how do we link those with the pipe segments and access them out in the field and blah blah blah. There's all these issues with that. And I want to try to figure out a way to deal with that, but I also want to figure out a way so that when I record a uh a pipe segment, like in two days, I want to know what the condition of that pipe segment is.

Brian Simmons: 36:49

So you're now spoiled by the violytics example, yeah, and you and you want to expand on that. Right. And and I I think it's it's good for you to be able to do that. I want it all, and I want it now to be demanding that way because because the technology exists, frankly. Yeah.

Marc Culver: 37:02

Yeah, you know, and and and so I'm excited at some point, hopefully in the next 12 months or so, we'll also jump into that world. Um, and then I can report back on that as well. Uh, talk to the nerds out there in the ether about this. Um but yeah, um, let's talk drones and AI. Okay. I mean, we we talked about this, you know, we're this might be a little there might be some redundant elements of this from Brian's previous podcast episode, but um where are you seeing the disruption uh in the industry with AI and and drones?

Brian Simmons: 37:39

So the the biggest and and probably I hate the term low-hanging fruit, but it the most accessible in the drone world are are again similar to uh you know dash cam or vehicle-mounted cameras, uh, things that we can see that a visual inspection tells us the condition. Uh, and so inspections, inspections of things like cell towers and cell antennas or power infrastructure are two of the ways that drone AI is being used heavily right now. Uh and specifically because that aerial perspective is important, you know, you're not gonna climb a cell tower just to take a picture right up there with your with your your phone or something. Um, you will have wasted the time climbing that tower when the drone could do that part for you. So there's the dangerous, yep. Right. And the repetitive is if I'm someone like Verizon, look at how many towers I am responsible for, you know, or or even in your case, you probably have some cell antennas strapped to your water towers. Yep. You know, and so inspecting those and looking for things like burnt connectors, you know, or even loose wires, um, things that have worn through, those are all conditions that can be identified. And this is where we go a little bit into like where are we running the AI? I don't want to spend too much time on this, but the edge compute of if if that train model exists on that drone or is accessible to that drone, we can now do that cell antenna inspection somewhere rural or wherever America, you know, the middle of Alaska. It doesn't matter if we don't have internet because that train model can actually be distilled and reside on that drone, and the drone can alert you to there is a problem here at in real time. And so you can be flying a drone, even in an automated mission, you know, doing an orbit around a cell tower. You could get a beep from the controller telling you that there's something that needs to be addressed and tell you the severity. Um, and so that that's just one way that we are combining the that's where the actually the term AI in the sky for this presentation came from was the inner intersection, the nexus of AI and drones. Yeah. Yeah.

Marc Culver: 39:40

And that that and that's really exciting. And it there are so many uh potential applications for this. You know, I know it's being used for tree identification, yes. Um, you know, actually identifying uh varieties of trees, but then also identifying uh disease trees. Yes. Um and again, you you talk about something that is Uh repetitive and time consuming as far as inspecting trees and and things like that. And to be able to do that on a regular basis, you know, from year to year, see the changes. Absolutely. Um is uh really exciting.

Brian Simmons: 40:14

Um, the progress of invasive species like emerald ash borer has become a public works problem. Yeah. And so now suddenly, you know, none of us are well, there might be foresters listening to this. I don't know, because there are definitely forestry nerds.

Marc Culver: 40:27

There are do you employ a couple forestry nerds on our podcast.

Brian Simmons: 40:30

Respect. Do you employ a forester? We do employ a forester. Excellent. Okay, yeah. Okay. We do. So you acknowledge that you have a bunch of trees in your care. Yeah. By having that person. Uh, but I'm gonna I'm just gonna say that not every agency and community has that. Right. And so then maybe I need to add to the four D's a 50, which is overwhelming. Yeah. Yeah. Well overwhelming, yeah. Yes, that when I when I think about a uh community that's probably small enough to not have a forester, right? Uh, you know, bringing that tool to them suddenly makes this data available.

Marc Culver: 41:01

Yeah. Uh, you know, one one of the there are lots of applications for uh pavement roadway pavement rating. There are a smaller number of, I would call them beta applications right now of trail and sidewalk condition ratings. And why why is there a smaller number, do you think? I think it's because they've been challenged with the mechanism of actually driving um those smaller facilities. I think I've solved that problem.

Brian Simmons: 41:36

I have I have a different idea.

Marc Culver: 41:37

Okay.

Brian Simmons: 41:38

I think there's the drone drone is one way. Yeah. Drone is one way. I bought an electric bicycle this year.

Marc Culver: 41:43

Yeah. There, there's there's actually I saw one of the one of the vendors, and I don't remember the name of the vendor now. One of the vendors I saw at PWX. They were what's really cool about PWX is they have these uh uh um like vendor showcases within the exhibit hall.

Brian Simmons: 41:56

Yes.

Marc Culver: 41:57

And so it's like their opportunity to have like a little mini presentation in a book there. Yeah, the new technology theater. It's a great name, even you would have thought I would have remembered that. Um, but I saw one, I I happen across one, and they were presenting a bicycle-mounted uh version of uh sidewalk and and trail uh condition. Very cool, yeah. And uh, so yeah, that is I think that's you know, and then there's maybe golf carts or you know, like a three-wheel, four-wheeler type of of thing or something, or toolcat uh type of device um vehicle that you could be on trails, but but it but now you're mixing that with trail users, you know, it's just a it's a different scenario. It is. Um and there's a little bit of danger in there. You mean danger in not running over pedestrians and bicyclists? That and also like let's let's say you put your staff on a bicycle. Sure. And now with a helmet, hopefully. With a helmet, of course, but they are exposed. Yes, you know, as they're crossing many intersections and and things like that. So there, you know, there's a little bit of a now, hopefully, the you know, it is safe to bicycle in your community. Um that that is that is a hope though, right? Yeah, not not an assumption in all communities, unfortunately. Um, but it just it is a complication because you have a different mix of traffic and users on those facilities. So I think what I like about the drone, and it's in it, and it's not a it doesn't solve the issue completely because I mean part of the beauty of some of our trails and sidewalks is that is they go into wooded areas. Yeah, you know, they're covered by trees, tree branches and drones, not friends, right? Right. So, you know, that that that's a complicated, and you don't want to have fly a drone at eight feet, you know.

Brian Simmons: 43:47

Well, that's a complication. You you could, but we're gonna be there for a little while. Yeah. If you think you're gonna capture all of your trails and sidewalk pavements that way, right, right.

Marc Culver: 43:55

So anyway, that that is a growing that is a needed application. Um, like I said, there's a few vendors out there, and I'm sure I'll probably get some feedback. Like, we're not beta. We know we we got this. Um, and good, please reach out, let's get touched. Show up in the comments. Um, let's do a product partner with Mark. Um but uh yeah, so what else? What else what else can we talk about with either drone applications or um AI in general?

Brian Simmons: 44:26

Well, so we ended this podcast with some comments about uh if you're not using AI sort of in your daily life, uh and and returning to that concept of uh I'm I'm gonna put AI to work for me, yep, right, so that I can do other things. I think we casually suggested that you should treat AI like an intern. You know, use it use it for a first draft. Yes, use it to edit for things like tone. Um, I I might have been known in my day to write a strongly worded email and then feed it to uh an AI and and tell it I need this message to be uh diplomatic and professional. Yeah. Uh and it's it's helped me out immensely. I see a good job once or twice, perhaps. Well, I man, I hope not, but um I I actually I I hope it helped me do a better job of communicating, is what I truly hope. Yeah.

Marc Culver: 45:13

Yeah, I uh I remember talking about that during the presentation, and and I I I think I said to the audience, like, if you are not using AI in some method, some manner in your job today, you're doing yourself a disservice.

Brian Simmons: 45:26

Yeah, you pretty much stood on your lawn and shook your fist at the whole audience with that.

Marc Culver: 45:30

Yeah. And now unfortunately, I am not using AI very much in my own profession, so I'm a bit of a hypocrite. Well, but but the problem with that is, and I don't want to throw my IT department under the bus, at least not fully, not right now. I mean, I'm happy to do so in other uh venues, perhaps, but um they are nervous about it. And and I was right and I will say we are public agencies. I I work for a public agency. So most of what you do is freedom of information acting requestable, correct? Exactly. And so we have some data practice issues, concerns. Where is it pulling this data from? What data is it pulling? Um what are we doing with the generated data?

Brian Simmons: 46:14

Um yeah, do we consider the generated data re you know, discoverable or somehow a matter of record, even if it's speculative? Right. Or even if it's wrong. Yeah.

Marc Culver: 46:25

And so these are these are some of the issues that we need to be uh honest with ourselves about and develop a policy around that. And so the that is my IT department is doing currently are I think they've kind of developed a draft policy around AI, and we need to kind of we're we're kind of talking about it, vetting it, and and we'll formally adopt something relatively soon. But what kind of drives me crazy is like we have co-pilot. There is co-pilot, there is a industry standard available out there tool that I can't use at all. Not right now. Okay. Um, and that's partly because of the license costs. We don't want to spend $300 per user. Copilot Pro is spicy, yeah, it is. Um and also not knowing how people are gonna use it and and some of those other questions about uh you know data prep privacy and data practices and and some things like that. So um I get it, but I'm just like chomping at the bit because I want to oh understand about I want to start using it more.

Brian Simmons: 47:37

Yeah, because you you you hear me talking about efficiencies games, right? And you're like, I want that. Exactly.

Marc Culver: 47:42

Yeah, exactly. I want to generate PowerPoint presentations using AI. I mean, I don't want to I want my present my presentations at work. Uh this is one thing I miss about Bolton and Mank is having uh a you know graphic designers and uh uh proposal specialists and marketing staff. Someone who looks at your stuff and goes, This is ugly, I will help you. Yeah, to like make all my stuff look good. Yeah. And and I want that at the city because I want my stuff to look good. I want to make a presentation to the council and I want it to look at I want to have cool graphics and um nice pictures and things like that. And I it just I don't have the time to spend six hours on a presentation, right?

Brian Simmons: 48:25

You know, to make it look pretty.

Marc Culver: 48:27

Well, because you could spend a lot of time on a presentation, you could, could so I'm looking forward to some tools, and you know, I Microsoft now, even PowerPoint has got some it must be like borderline AI type of stuff where like if you're working on a slide, it will show you like options. Can I give you a word?

Brian Simmons: 48:48

Give me a word. That is agentic AI. Agentic. Yep, that is it. That is Microsoft Copilot operating operating as an agent inside of PowerPoint, Word, Excel, etc. Very nice. Yeah. So if you were signed into something like Copilot Pro, that it it's just a different way to interface with it. Yeah, but it's it's designed to be native with the document itself. Yeah, so it's over here going, you just pasted an image. Would you like that centered in your slide? And it luckily it doesn't say something like one of our wives would say, which was would you like that centered in the slide? Knucklehead. Yeah, because this is the ugliest thing I've ever seen in my life. You want me to fix it for you? Right. You're not allowed in the paint aisle at Home Depot with me anymore. Um, but you know, you dress yourself today. Right.

Marc Culver: 49:32

Uh I love I love that word agentic, um, and it actually gets me thinking about kind of the next uh topic that we'll maybe end up closing with here. But um like the future of AI, at least in my eyes, like how I'm gonna use AI in my job, in my profession, uh, and I expect my uh employees to use it. You know, we we are, as I mentioned earlier, we're in the process of implementing asset management. We're we're implementing OpenGov, formerly Carter Graph. OpenGov's enterprise asset management system. Like they got rid of the Coca-Cola of asset management.

Brian Simmons: 50:08

Carter Graph did have a ring to it. They did.

Marc Culver: 50:11

It did ever like everybody knew what it was.

Brian Simmons: 50:13

Yeah, someone said Carter Graph and I had fond memories of Dubuque Iowa.

Marc Culver: 50:16

Yeah, yeah. So anyway, I'll uh Mr. Onkin just waved us, Mark Ankin. Um I the open gov people can uh be mad at me uh later when they listen to this, but um anyway, we are just glad you're paying your bills, that's all exactly, exactly, and it's not a small bill. Um we are uh implementing asset management, open gov, and I'm really excited about it. But it is like, and my staff, I think, is intimidated by right now because it's like okay, we're gonna have all this stuff in here. We're gonna have all this data, we're all this information, we're gonna have all these assets, you know. We're we're trying to keep up with you know inputting our maintenance um data into it, figuring out how do we assign condition ratings to all this stuff.

Brian Simmons: 51:07

At what point does the time spent asset managing no longer result in efficiencies? Right. Like are we just cartographing for the sake of cartograph, or are we doing open gov because it brings us efficiencies? Is that is that a good way to summarize that?

Marc Culver: 51:22

It it is, yes. And and my concern is not so much like because I still think even if we didn't really do anything with the data, you know, from an analytical perspective, and it was just there for us to reference it, it still has value. It still has value, you know, because otherwise, that is institutional knowledge that is in the head of my maintenance staff that eventually goes away. Um, and so there's still value in capturing that information and and putting in a repository that we can access. But what I'm really looking forward to, and this gets into the this agenc uh thing, is I'm looking forward to, and I know they're working on this, I know they've made some uh acquisitions recently, I believe, that is going to help them build some AI elements, some agenc elements into their system that will give us like outputs that's that's going to give us reports, it's going to give us uh stuff that we don't even know what to ask for.

Brian Simmons: 52:18

So instead of spending time trying to teach copilot how to work with the data that exists in your asset management program, your asset management program will inherently ask you, what would you like to do with the data that I am holding for you?

Marc Culver: 52:30

Yeah, is that that's your or or it will even give us uh hey, did did you know this? Here's a dashboard your infrastructure. Here's a dashboard. Look at this red flag you weren't thinking about. Exactly. You know, I can say build me a dashboard for this thing, and it will build me a dashboard.

Brian Simmons: 52:46

I feel like I want to say that all the time. Like a dash dashboard is always just show me the data. Yeah. Show me the data. What am I what am I missing? Right. How am I not interpreting this? Yeah. What would what one yeah, exactly.

Marc Culver: 52:58

What what am I not doing that I should be doing? For sure. What what question what question should I be asking? Yeah. What do I not know that I don't know?

Brian Simmons: 53:05

Is that is that like the smartest thing you can do? Right, is just say what what am I missing? Right. Right.

Marc Culver: 53:11

So anyway, I'm really excited about the the premise of that. I don't know when we'll see that kind of roll out and I'll be able to take advantage of it. But um I I mean I consider myself a pretty seasoned public works professional. I'm a nerd. Yes. I like to nerd out. Um, but part of the part of the problem with being a true public works professional, particularly for a you know, mid-sized city, is you have to be an expert on so many things. Yeah, you do. You know, I mean I you have to be adept, right? In very many places. Yes. You know, I I I I could I call myself uh I joke about it, I call myself a recovering traffic engineer because I was a traffic engineer. I was an expert at traffic. Once upon a time. All right. I could bore you to tears and talk about cycle times and splits and and offsets and stuff like that. And I think when I got to that part on the PE test, I just made Cassiopeia. Right. Well, I did that for uh Geotext. See, there you go. There you go. And and that's great, except now I have to talk to my council and my residents and uh engineers and consultants and business people about sanitary sewers for sure and flow and drop manholes and lining and uh water pressures and um you know water towers and pump stations and storm water, uh you know, infiltration rates and uh PFOS and softening and uh recycling and uh pavements. Manganese. I mean not manganese. Yeah, we got manganese, but you have a manganese treatment plant. Um, I mean, just so many things so many that are in our control and we need to know things about. And so when I get something like an asset management system, I don't know what to ask it. Like, all right, here's my sanitary sewer system, besides trying to figure out a condition of my pipes. What else should I be doing with this data to manage it? Right, to manage these assets for sure. And yes, we have our utilities managers and you know our division managers, we have our experts within our our uh departments and such, but still we we gotta be able to talk to talk.

Brian Simmons: 55:34

Yeah and make decisions, and make decisions, right? Not not just not just sit back and and watch things happen. Yep. You are on an annual basis being asked to allocate money based on some of those things.

Marc Culver: 55:46

Yeah. So in any event, um I am just like super eager and uh can't wait. And like I said earlier, chomping at the bit, to use AI in every possible way that I can to make my job easier, more efficient, to to make me smarter, um, and sound smarter. Um to do your laundry? Yeah, exactly. To do my laundry, yeah. So I can make art.

Brian Simmons: 56:15

Absolutely. Or at the very least, smooth roads. Which is art unto itself, I hope, to a traffic engineer.

Marc Culver: 56:23

Right. No, well said, well said. Anything you want to close with, Brian, on this on this exciting AI topic? Uh I you know what are you looking for?

Brian Simmons: 56:30

Like what do you see on the horizon that you're just like super juiced about? So so the thing I'm excited about is what comes next after the agentic piece? Yeah. At what point can I start connecting pieces of AI? Or can I start allowing AI with my permission, of course, because we are all about AI for good, but allowing AI to start to make some decisions. You know, there are some systems that we are working with right now that will, you know, input weather data uh and give us uh, you know, plot that against proposed or uh predicted storage in a pond system and tell us, you know, there's this huge weather system incoming, you are going to be under capacity, and then start making some decisions about what to do with it. That that next step beyond the agentic AI, and I don't have a good name for it. I'm sorry. I'm sure there is one out of there in the AI world. But when we start allowing the the AI to start to at least recommend some decisions, if not execute on them, yep, to start to do some of those things. That's that's where I think we're gonna see the real excitement.

Marc Culver: 57:27

Yeah. We uh in Brooklyn Park, we have 700 acres of potato fields that we will be developing over the next god only knows how long. And we're excited about it. And we're super excited about it. We are, and and part of that area is this biotech area, and and your firm is helping us put together a plan and an AUAR and uh you know an infrastructure plan for that. And one of the things that I'm working with uh our dear friend Tim Olsen on right now uh at Bolton Mank. Uh love Tim. Um big fan. Tim Tim was my office mate, and I keep we keep threatening to do a podcast, him and I, and we gotta actually pull the trigger on that. I would like to be the live audience when that happens. Yeah, you you can heckle us. Um but Tim was my office neighbor for for a while while I was at Bolton Mank, but he's a genius. He's a he's a he's a stormwater guru, him and Ryan Johnson from Rose Ope. But in any event, we are we are trying to design, we have this goal of designing a regional stormwater system in this biotech area rather than having all of these parcels do their own stormwater on site. You know, let's develop a regional stormwater system that looks good, yeah, that we can maintain easier, but also we are super blessed in those areas with sandy soils. We have super high infiltration rates. So it's just gonna go away at some point. So we're we want to work with the watershed, we want to push the boundaries because usually there's a cap on how much how fast we can infiltrate. Yeah. And like the ultimate goal is to have like a zero just discharge stormwater system. I think that's what the watershed districts would love for you to have. Right. They they're not allowed to make that rule. Right. But so we're gonna try to figure out how we're gonna push some of the limits on on some of this the infiltration rates and th and that. Um, and how about how about the ways that we're gonna push limits on that together?

Brian Simmons: 59:07

How are we gonna push those limits together? Instrumentation. I'm sorry, what instrumentation?

Marc Culver: 59:12

Instrumentation, yes, yeah. So we're good. So that's part of this is measuring this and and using applications like you're talking about. Let's move water from this point to that point so we have space in here for the storm that's coming, so then we can infiltrate all that over time and keep it from blowing out because we, you know, it's just coming in faster than we can infiltrate it. So in any event, uh I'm just super excited about doing stuff like that. I'm super excited about putting sensors on our water mains and our sanitary sewer pipes and our stormwater pipes, um, as we build all this new infrastructure out there to monitor things and low data, level data, all of those things. We're gonna we're gonna we're gonna make that portion of your city as smart as possible, I hope. Amen. It's gonna be a biotech area and a smart infrastructure area. I just I just thought of that. We're gonna create a smart infrastructure district. It's gonna be a SIF district. Yeah. Let's go with that. Let's go with that. All right. This we've probably been talking uh long enough here. You're gonna edit this. Uh maybe, maybe not. Okay. Um, we are gonna, you know, there'll be a little bit of editing to get the good audio from the podcast in here, and maybe we'll put a little bit more in there. Um, but uh thanks for joining us today. Thank you for joining me again, Mr. Simmons.

Brian Simmons: 1:00:21

It's my consummate pleasure, thank you.

Marc Culver: 1:00:22

Yeah, and I'm sure Brian and I will be talking again about nerding out about some other topic uh at a future episode someday as well. Hey, thank you for joining us today. Um, I'm sorry that we haven't been dropping and I don't know who we is, by the way. I keep saying we when I talk about this podcast. It's me, it's a one-man show right now. Um, but I'm sorry that I have not been dropping episodes uh on a more regular basis. I'm gonna work on that, but I do have a full-time job now that this podcast is not a formal uh official part of, but and it's a job that I'm very passionate about. So I spent a lot of time on it. But I'm passionate about this too. So I will get back to this. We'll be dropping some more episodes. Um, thank you as always for listening. Um, please look for our posts on LinkedIn, share that, like it, make a comment on there because uh, you know, we we want to grow our fan base and uh see more people out there. So thanks again, and we will see you next time. Nerds out. Nerds out