What happens when AI tracks every resource in your city?
Unlike most cities, which are burdened with aging infrastructure, McCord Development had the rare opportunity to build a city from the ground up. But what does it really mean to create a smart city? In this episode of Your AI Injection, Deep Dhillon speaks with Ashwin Chandran about the AI sensors used to track water consumption, optimize energy distribution, and rethink urban planning like parking lots. The two dive into the challenges of integrating AI into city infrastructure, the unexpected ways data is shaping development, and whether AI-managed cities could be the future of urban life.
Check out more about Ashwin here: https://www.linkedin.com/in/ashwinchandran/
and McCord Development here: https://www.linkedin.com/company/mccord-development-inc/
Check out some of our related content here:
Get Your AI Injection on the Go:
xyonix partners
At Xyonix, we empower consultancies to deliver powerful AI solutions without the heavy lifting of building an in-house team, infusing your proposals with high-impact, transformative ideas. Learn more about our Partner Program, the ultimate way to ignite new client excitement and drive lasting growth.
[Automated Transcript]
Ashwin: cities are losing, big cities are losing millions of gallons a day and nobody bats an eye, right? So it was a definite change in paradigm for me when I'm looking at oil. I was like, how can you just, you know, lose water and not know where it went?
Deep: Well, especially given, you know, the changing climate and the fact that we're, water is going to become a lot more expensive as we're increasing desalination and, and having to bring water in from the oceans.
CHECK OUT SOME OF OUR POPULAR PODCAST EPISODES:
Ashwin: we lost 3 million gallons of water over two days, two years ago. And that's one of the things that drove this. Okay, we need to build something because we don't. I mean, even 3 million gallons of water is few thousand dollars.
So it's not really like you're talking about millions of dollars, right? And in a city, that's not a big expense compared to what you're doing, but losing 3 million gallons in two days. If you just think about that number, that's huge, right? it doesn't compute for me.
Deep: Hello, I'm Deep Dhillon, your host, and today on your AI Injection, I'm joined by Ashwin Chandran, [00:01:00] Director of Technology Innovation at McCord Development and co founder of HEX20 and Datron. Ashwin holds an MS in Electrical Engineering from Virginia Tech and has a deep passion for leveraging AI in myriad innovative ways, from designing and developing cutting edge solutions for a 4, 300 acre site.
Smart city near Houston to pioneering new approaches to satellite design. Ashwin's at the forefront of applying AI to complex development challenges. Today, we're going to dig in to how AI is turning once futuristic ideas into present day realities. Ashwin, thanks so much for coming on the show.
Xyonix customers:
Ashwin: Thank you, Deep
appreciate you having me here.
Deep: Awesome. So maybe let's get started. Tell us what does it mean to build a smart city? And what kind of a role does AI play in that and maybe focus on or what happens currently without your solution and, how that's different once your solution is in place.
Ashwin: That's, you know, great place to start. And smart city [00:02:00] is such a nebulous term. You talk to different people, in these technology walls, doing smart city related, technology development and implementation, and you get a lot of different answers, what it means to them, you know, we are developing, like you said, in my introduction of big city, all greenfield completely from the ground up, About 4300 acres, which is about quarter of the size of Manhattan, right outside of Houston city limits, right?
So it's not trying to shoehorn technology into an existing city, trying to build something that doesn't exist today. It's going to be literally for tomorrow, but because it does not hear today completely. So it's all new.
Deep: So is it a, is it a full city or is it like a suburban development or something like,
Ashwin: yeah, it is a private public partnership, right?
So I'm on the private side of the build, which means that we own the land where the third [00:03:00] largest landowner in the second biggest county in the U. S. So these type of greenfield developments don't happen often, especially in the U. S it's a very big opportunity because it's not a, cookie cutter development.
it's, commercial, industrial. everything from, you know, multifamily building, no single family, but multifamily, commercial, industrial food and beverage. Everything is here. So it's it is a city that has to support, the citizens that are around it as well as we'll be working and living in this area.
So what a smart city means for us might be very different from what it means for, an established city.
Deep: And are we talking about just infrastructure, like electricity, water, housing, and buildings, or are we also talking about the, government infrastructure as well and the governing infrastructure?
Ashwin: that's a great question. So I mentioned this in the beginning, but I didn't elaborate. It is a private public partnership. I only explained the private side. The public side is [00:04:00] the generation Park Management Board, which is, the public entity that is responsible for tax collection, bond issuance, all of that.
So we build the infrastructure, whether it's roads, bridges, parking garages, parks, whatever it is, and then the, management, board takes over or the public entity. Okay. Entity takes over the infrastructure and pays back with bonds or whatever, mechanism there is, right? So it's it is truly a private public partnership.
Deep: I'm just curious before we get into it how does this come about because typically this is not what happens, right? Typically, there's just a county with maybe unincorporated lands and people just start building there and eventually there's a bunch of them and then, you know, there's a somehow there's a post office and it turns into something what happened in Texas outside of Houston to, get something so so planned.
Ashwin: I think it's Not that common, but it is not that rare in Texas or right outside of [00:05:00] Houston. If you know the suburbs of Houston, which you might not, the Woodlands is one of those established neighborhoods, that is similar in structure, which was a private public partnership with a private entity holding a lot of land and trying to develop it.
And a public entity that that's responsible for the governance and then takes over the public infrastructure once it's developed. So it's a similar structure to that that we've established here I think it's one of the really productive ways of developing a city It short circuits a lot of the decision making process in a, in a normal city. Because it's being developed, the private builder has the ability to invest in technology, and promote solutions that the board can then approve or, you know, say, okay, this doesn't make sense.
So that that decision making process is pretty quick, pretty fast, we can pilot technology throughout value and then scale. based on its [00:06:00] efficacy rate. it is a short of a test bet that we can establish for tech when it comes to technology and decision making is quicker than in a normal city.
Deep: It's probably not. That different from like a suburban development scenario from a like how to get it done standpoint like you interact with the county or whoever has jurisdiction there. maybe the difference here. Isn't that you're not just putting in housing or putting in one type of, construction. You're kind of doing a more mixed use scenario. And the scale, the scale is probably the differing factor. Would you agree with that or
Ashwin: the scale? And I think one of the more important differences might be incentives, right?
What is the county trying to do? And what's the private builder trying to do is often not aligned. So I think there's a lot of friction there sometimes in a, in a normal setting, uh, which you have to overcome. And I don't think we have a lot of that friction because we're aligned in a lot of our objectives.
when it comes to the development here, usually when a [00:07:00] land about, you know, the size of land exists, it's very much in the real estate developers interest. Okay. To build quickly and get out.
Deep: Right.
Ashwin: Right. Build quickly, sell and get out. But that's not the vision. The vision is to really, we, this has been going on for about 13, 15 years somewhere now.
We're only like 10 percent built out. So we're taking time. It's project by project. the vision is to really build this out into something good. And that stands the test of time, not just to build a small development, right?
Deep: Does the does your firm have a vision to? make money off of capital construction, like building buildings only.
Or is there some kind of managerial or like recurring revenue scenario that they're thinking through?
Ashwin: So if you look at it, the biggest asset that We hold as a land, right? And the objective, really, I mean, if you look at it from a financial standpoint is to maximize the value of the [00:08:00] land, that's ultimately the financial objective, right?
And that does not come with a short sighted vision of build and sell, but it comes from, hey, how does this land gets valued more over the years is by attracting the right kind of commercial development into this area. Right. Which might be pharma manufacturing, which might be high tax base coming into the area is what is effectively going to increase the value of the land over time, right?
And how that is maximized is really the objective. And that, that's where, you know, from a private builder financial standpoint, that's where the incentives are. Bringing in that that high tax base is effectively what the public entity wants as well.
Right. Right. How do we bring in the bring high tax base into this area, which is east side of Texas east side of Houston. If you know, this is probably the only pocket that's left to be developed. So how do we bring the right tax base here [00:09:00] and that, you know, uh, increases the tax value of the tax collection for the public entity and increases the land, value for us,
Deep: is the, the way that your firm thinks about the value, because presumably once you sell a house, you sell, you've checked and you've cashed in the value of the land at that point in time. Right. But is the idea. That it's like a long time arc project. And so the value gets kind of appreciated over time.
And then there will be properties being sold later on, like, five, 10, 20 years down the line, or is the idea. to realize the value up front and then be out.
Ashwin: I'll go back to what you said with a slight modification. When you sell the land, , the value of that particular plot is realized.
Deep: Uh huh.
Ashwin: But depending on who you're selling and what is coming in, the value of the, if you look at it as an entire, uh, 300 acres, That is still to be ascertained. You could increase the land [00:10:00] value by bringing in a particular type of project, right? So even though you've made the sale and realized the value of one portion of it, it could still affect the value of the surrounding area quite a bit.
So that is also a big factor into what comes in here. What are the projects we're going to target? Who should we bring in? So if you look at the strategy here, one of the big areas or one of the big industry sectors that we're trying to target to bring in is pharma manufacturing.
We've got a big med center in southwestern Houston's the biggest in the world. It's got pharma research. And clinical studies all going on there, but Houston doesn't do a lot of pharma manufacturing. So how do we transition that research and clinical studies into manufacturing and provide the workforce that we invested along with Nyberg, which is an, Irish.
training workforce training, group that does a lot of bio pharma manufacturing to [00:11:00] collaborate with San Jack, which is our local community college to start a facility right here for bio pharma, workforce training. Right. So that is, is essentially going to provide a lot of workforce for the companies that we want to try to come in and build their manufacturing facilities right here.
So the groundbreaking for this has already happened, and that facility is coming up right now.
Deep: Right.
Ashwin: I mean, one of the things that It's a long term plan.
Deep: One of the things that's unique about what you're telling me is these feel like the realms of government, right?
Like cities think about how to attract pharma manufacturing, usually private companies do not. Actually maybe from both vantages, cause you could have just focused on.
Building out a bunch of suburban houses and sold them and been done with it, right? Like that could have been
Ashwin: absolutely
Deep: approach. It could have viable
Ashwin: strategy.
Deep: Well, I mean, it's prolific in the US and Canada, right? Like this happens everywhere. So, so what were [00:12:00] the forces that led you to think? I want to do attract something like pharma manufacturing.
Why not just set aside that land the way the city does put up zoning or something around it, you know, work with the city to put in the zoning and then just leave and just be like, well, whatever it's up to someone else to do it. Yeah.
Ashwin: Well, one thing Houston has no zoning. I think this will blow a lot of people's minds.
Houston has absolutely no zoning. You'll find everything. in neighborhoods too. Uh, Houston, small, those cities. . Uhhuh. Yeah. No zoning. No zoning. So, so given that, if you look at this area, there is enough single family housing around going on, being built. Building single family housing, and getting out is definitely a strategy, a very popular strategy, like you said, and it just did not fit with the ethos.
It's a very family oriented company. I mean, it was started by our current president, Ryan McCrory, his [00:13:00] dad. I was the guy who built the company and Ryan's got this vision of, we are stewards of this land and we're not going to just build and get out. We're going to hold us and build us in a way that is, they're very much, part of Houston.
So take
Deep: pride in
Ashwin: being from Houston and, giving back to the area. So giving back to the area is attracting those right kind of assets here that is going to continue the development of the city as a whole. It's not just Generation Park. What we're trying to implement.
Deep: It sounds sort of like Vulcan, like Paul Allen's group here in Seattle.
they had like deep ties to the city. Paul has had lived here, you know, much of his life. Um, and they sort of sat down with South Lake Union and decided that they really wanted to, invest specifically in like pharma manufacturing so they ended up redeveloping that entire chunk of the city, a ton of, civic, kind of cooperation, but, okay, so that, paints the picture [00:14:00] about, you know, you've got a big chunk of land, you're in a city that has no zoning, I don't know what to do with that information yet, and then you have Some, civic minded, folks on, uh, both sides of the fence thinking bigger than, residential housing or something thinking kind of more, integrated and, and then honing in on specific industries.
So let's talk about why is smart city? Is it just because it's 2024 and it's new and, there's an opportunity here or like, and what level are we talking about? , can you lead a group that focuses on AI in this context? what does that even mean? Are you talking about infrastructure?
Are you talking about AI for like supply chains to get your, you know, construction costs down. Like, what are we talking about?
Ashwin: I direct the technology team here, but technology is for us. It's not just AI, right? But AI is a part of it. So for us, really, why a smart city?
And I don't think I answered your first question, right? What does it mean to be a smart city? We got, , down this [00:15:00] rabbit hole of what we're doing. Welcome to your
Deep: AI injection. It's just the way my brain works. I just grab something and start digging down and eventually we'll hit some bread.
Ashwin: No, that's perfect.
I love that.
I love that. But, I worked for oil and gas for a long time, right? So I saw this role pop up and I was like, well, that's interesting. Let me talk and find out what's going on here because I had no clue what was going on here. Right? I've been here about three years now.
So talking about this, you know, understanding what's going on, what, the concept was, why do we, they need to typically look at a real estate, um, development for them. They don't have a director of technology generation. Most of the time they don't. So, what's what's going on here? Why do they need 1?
And what's the idea of having 1 somebody on board to do that? You know, the, the concept really is to use technology wherever possible to 1, you know, remove friction from daily interactions of [00:16:00] people with the city. It offers services, increase quality of life for the users, offer a secure environment where possible, take care of the natural resources with technology, and then another key principle is to use technology to build and design and plan the city.
That's one of the things that might not be applicable for a lot of people because they're working in established cities, we are effectively building something totally new. So how does technology play into what decisions can be driven by data so that you're not doing something wrong?
You're doing the right thing. What do you build next? I get down these rabbit holes to to so stop me, you know, I'll give you a good example. The 1st sensor to be deployed here in Gerson Park was a parking to like a car, a car counter, right? And pneumatic to.
So we built a big, FMC Technologies is one of the anchor tenants. They built their, U. S. headquarters [00:17:00] here, and they, that construction started about 12 years ago and we built a big park right in front of their headquarters. Public park with a big water, structure, you know, all that's beautiful. And there was a small parking lot that was gravel parking lot and driving past it saw cars there.
In the morning, evening, of course, we're there and we're curious, like, who's coming here? Why are they coming here? How many cars are here? So, parking tube was kind of the first sensor to be deployed to figure it out. And that really drove some of the decisions for us. Okay. Because what we figured out was, you know, 300 cars on average are parking there every week.
And they're coming from the neighborhoods around us. Because there's not enough green space here. So that drove a big decision. Okay, green space, parks and trails are going to be a big part of our ethos and we're building a four mile walking [00:18:00] trail around the west side of the city. We're building trails that connect our generation part to all the trails that are in Houston.
So this, really drove some of these decisions. So data, you know, when we talk about making decisions based on data, okay, this was one of the first things that drove that decision. And that's where, okay, we need to start deploying technology to understand more about how our assets are being utilized.
Deep: Well, let's talk about that. I feel like, sensors and, and the data pipeline can be a, an interesting entry point into the AI. So do you guys think about data collection? Like, what are the sources of the data? To what extent are you thinking about sensors that a city has? Like, so typically Cities have you mentioned, you know, car counters.
That's certainly one. usually there's a 911 emergency response system. That data gets kind of fed up and, into the cloud and made accessible via api's via company that I worked for few years ago. 311 is another one. You know, [00:19:00] people, there's a lot of civic apps that people build. They, you know, shoot a photo of a broken window.
It winds up on the 311 system if there's a car sitting unwanted in front of somebody's house, it usually winds up getting reported via 3 1 1 system. and then there's depending on the vertical, there's all kinds of stuff in, in GovTech. If you're talking about financial stuff, everything from all the checks, the city rights to all of, you know, including like the payroll of of the staff themselves.
to how are they spending their money broken down by topic. So how do you, how are you guys thinking about sensors different from how, let's say, an established city might think about it,
Ashwin: right? So I think, you know, one of the things that go into and one of the key roles that I was initially assigned with was, To build a smart city, right?
What, what are the different enabling pieces? And one of the enabling pieces really is sensors and the way they communicate. What is that communication channel? What are the sensors we want to deploy? And where do we deploy them? Uh, and how do we [00:20:00] get that data aggregated, collected, tagged, cleaned, all that stuff, right?
So we've established a LoRaWAN network. here in the city as a pilot to start deploying a lot of sensors, whether it's air quality, those kind of sensors in internal or external, internal to buildings or external to the city. Now, you have to understand we are only like 10 percent built out.
So a lot of the sensors that you mentioned in a built environment and a built city might not apply to us correctly.
Deep: Right?
Ashwin: Because we're still building and trying to look forward. for us, the biggest verticals that I'm focused on, I'm very tiny team, right? So, what I'm focused on right now are, one's water, and two is, parking. Water and parking are kind of the two main verticals that I split my time between. mainly because. Water, we started off with a complete new infrastructure, new water, new, Pipes, new water plants, everything, new spot meters, including our distribution meters.
But when we talk to, I don't know if you know much [00:21:00] about how water operations are set up in a state like Texas, in small, small municipal districts like ours is the district does not have any employees to manage water. So they have a private company come in and operate their assets. Right. So we talked to the interview private companies to come in and operate our assets.
And we say, Hey, everything is new here. We want to know 100 percent of where our water is going, not bill for it, but be accountable for it. Right.
Deep: Let's start, let's start at the beginning of this problem because I, I like this thread and so you guys have blank land. You need water. Where does the input come from?
Are you pulling it from the ground? Is it, are you, are you tapped into the, the city's supply? Like, where do you even get your water input from?
Ashwin: to both of those, right? We have wells that will provide our water. We also get water from the city of Houston. Okay, on a take or pay contract. And we have also built [00:22:00] reservoirs now that will supply our non potable water.
Right, that will reclaim rainwater and will supply our, our non potable because earlier it used to be all the water, whether it's potable or non potable is from either the well or the city of Houston. Right.
Deep: Got it. First thing you got to do is build some kind of entity.
To build the well or get somebody to build the wells and then somebody to manage the wells and the intake from the city. And then you had to build some, somebody had to either acquire or build some software to manage how much is being pulled out of the ground and how much is being, pulled from the city.
And then you had to send it somewhere. So you had to build a, some plumbing infrastructure citywide. And then you had to take that into all of your buildings. So what sensors have we talking about so far? I imagine we're talking about inflow and outflow and maybe time of day. what else?
Right.
Ashwin: flow meters everywhere. there is a water plant. That's one of the things you miss. There's a water plant that, that processes the water, right? So water quality measurement, those plans. currently are not, instrumented, [00:23:00] but they have pumps,
so this is where the operator comes in. The operator is a private entity who manages the infrastructure. Whether it's plans, whether it's all the meters that are here, even though the district owns, it's managed and operated by a private entity,
Deep: right?
Ashwin: So, and often these incentives don't align private operators have definite.
The district has another incentive when it comes to managing their assets and how some of these are not aligned. And that's usually where we run into issues when it comes to water. Plus water is cheap, but it's the cheapest it'll ever be in our lifetimes. It's all only going to be more expensive, right?
So,
Deep: , let's see, so you've got the flow meters tells you like how much water you're pulling in, probably, you know, the cost of the water you're getting from the city and maybe that's dependent on how much you use and does it maybe time of day or something?
Ashwin: No, it's a take or pay.
So the contracts with the cities are usually, you're paying for [00:24:00] a certain amount of water, whether you take it or not. it's there, you better use it.
Deep: And then presumably the stuff you get from the city you don't need to treat, but the stuff coming from the ground you treat.
Ashwin: No, we treat everything, right?
So the plant is, the plant treats everything. So basically the architecture is that the city and the well will go into a holding tank. And from the tank, it goes into the plant.
Deep: Okay. and now you've got your water and then you have a demand that fluctuates based on. Seasonality effects based on the population's daily habits and weekly habits, and summertime, maybe people are filling up pools, wintertime, maybe the consumption's less.
I imagine, like, if we start talking about where AI is helpful, is it helpful to predict and forecast you know, how much water you need from the city and how much water you need from the ground,
Ashwin: Prediction and forecasting is one. Yes, absolutely. The other thing is understanding. we talked about building it correctly, right?
How do you drive decisions based on some of this? Some of the dark data [00:25:00] that water provides is really interesting. It's a big present sensor, know, after COVID We knew which buildings were more occupied than others, right? If you look at office buildings. things like that can actually start driving more intelligence into how assets are being utilized.
Now, we have, we built one water plant, but we're definitely going to have to build another water plant. Knowing our utilization, knowing how much overhead we have, knowing what can we support. Like, we have a surf park coming in. That's got a huge capacity requirement.
Can we support it? What's a surf park?
Deep: People wear surfboards and they, and they surf over it. Yeah, yeah, yeah. Like, like
Ashwin: artificial, artificial waves.
Deep: Uh huh. There are a
Ashwin: few, few around the U. S. This is probably, this was designed to be one of the biggest or something. I don't know exactly the details, but So, you know, that's a huge demand on the water plant and can you support it with the existing plant can be.
So what are the things, capacity planning is a big [00:26:00] piece where this helps because we know actual usage. There's a planning part of it where you have to use certain models that are prescribed by the governing authority. But then you can go back and tell the authority.
Hey, these models don't really make sense. Here's our actual usage. We can model it based on our actual usage. And here's our overhead based on what we see the ground right now. So we have the ability to go back. And then do that. And where I has made some actual real impact on the ground is anomaly detection.
And going to our operator and telling them, because we know more about the city and its water usage today than our operator does. So we can tell the operator, you have a leak in this location, go fix it. Ideally, this should be automated. Right? Ideally.
Deep: Uh huh. and this might not be like a residential leak.
This is like some kind of, something in the street moving, uh,
Ashwin: exactly. I came from oil and gas, right? if [00:27:00] oil is, if we can, we don't account for every drop that's being produced, there's an issue, but water is not like that. Water, you know, it. Take some. Yeah, nobody cares as
Deep: much because it's water.
Nobody
Ashwin: cares. Yeah, and cities are losing, big cities are losing millions of gallons a day and nobody bats an eye, right? So it was a definite change in paradigm for me when I'm looking at oil. I was like, how can you just, you know, lose water and not know where it went?
Deep: Well, especially given, the changing climate and the fact that we're, water is going to become a lot more expensive as we're increasing desalination and, and having to bring water in from the oceans.
So let's talk about that specific case because I find that interesting. So like. what signals are used to predict, a leak. So you have a known input flow going in and you have a known consumption flow cause you have flow meters on the consumption side too.
Ashwin: And then multiple ways, right? Once based on flow, like you said, you know, flow should definitely give you and consumption meters. You can put multiple [00:28:00] signals there you know, depending on what type of entity it is. If you see continuous flow, right on consumption site, you probably do have a leak somewhere. Now, that's not really the best indicator when it comes to industrial meters, but then anomalies really based on Historic usage, seasonal usage, all of that. So we've, we have algorithms where we can do by individual meters or by hierarchy, we can set, alerts to, you know, alert us when these anomalies are detected based on historic usage patterns.
So that, that's one way. Now, the holy grail for us would be pressure maps. If we have pressure, which we're starting to get, but if we start getting pressure at input and output and along the way somewhere, then now you start building really good pressure distribution on your water distribution system.
Now, you know, even tinier leaks that are happening within your distribution set system with better localization, right? Yeah. So where are those leaks on?
Deep: Let's talk a little bit about [00:29:00] the flow meters, those sensors. So you have these flow meters. how are you centralizing that data, is this via like citywide Wi Fi or something do the pipes that are underground all carry power with them to power the, the flow meters or?
Ashwin: No,
Deep: are they just powered by the flow itself?
Ashwin: No, the flow meters are lTE communicating battery powered flow meters, right? So flow meters are not underground. Flow meters will be at junction boxes above ground.
Deep: Oh, okay. But somewhere there's a turbine. Spinning.
Ashwin: Yeah, it could be either ultrasound flow meters or or these spinning jet flow meters, right?
Deep: Uh,
Ashwin: but the bigger so so they come in there. they're communicating the data back. We actually have to build a service, right? We have to build a service to aggregate the data and build analytics on top of it.
Because, being a city, I really want to consume these services and products. I want to be the end user. But we go into a problem thinking that we're gonna, okay, look at what's available and [00:30:00] figure out what we want to use, but then end up building some of these ourselves just because there are a lot of gaps and it's not there.
So, we ended up building a service, which actually, right now, our operator is talking to us to start commercializing and start piloting the technology in other districts. Because there's value, so that's basically, smart meters everywhere. We're just using an API to bring the data into our service.
And once you get that flow data along with some temperature and some pressure data, we're starting to build some really good Insights about our water distribution system. So we're calling it a digital twin for our water distribution, right? It's an IoT platform.
Deep: walk me through so I imagine there's a few different scenarios that could have played out.
One could have played out like, oh, we're not going to have flow meter or anything, but I imagine cities probably put flow meters and pressure sensors out anyway. I'm totally naive to the domain, but I'm just going to throw that out there and correct me where I'm wrong. on the other extreme, you might be like [00:31:00] plastering sensors everywhere and being really aggressive about the, the data collection.
, and then the optimal cost per spend for value out might be somewhere in between. Walk me through like, how do you justify the added expense, is it just something that you guys are doing largely on faith because you want to be a smart city, or do you have to actually go through your team and like justify?
why you're, pushing for X number of sensing and what the actual value is, a. k. a. you know, water savings costs. And you have to do a lot of modeling there to like, get that justification. Walk us through your world a little bit. Like, how did you even get to the point?
Where you have these sensors and how do you determine how many to put out and all that?
Ashwin: So flow meters are common. Like you said, flow meters, every city will have by design, they'll have them because they have to measure. But what type of flow meters they have is different, right? older cities, older infrastructure all have non smart meters.
If somebody has to go out [00:32:00] there to read it. and then there are, meters where you can drive by and read. they have a handheld and drive by these meters and they get readings. Thanks. And then there are meters that are like ours, which is truly smart and can communicate back to a central server.
And then, you know, that data is there. But how many of those are there and how this data is collected and what is being done with the data not at all standard, right? Because, Okay. You know, I've talked with cities where there are smart readers in these pockets, they get the data, but they still don't use it.
At all or anything other than billing, So what we do with that data is what drives the rationale behind how many we need and how we can rationalize some of this cost as well, When we decided to build a system, all of the meters, the big distribution meters were going to be smart meters.
That was a decision made at the front end. Okay, we're going to build something new. Let's build something that is. going to service and in the future. So the district don't bore the cost for those [00:33:00] meters on just a concept of this is the right thing to do without a lot of justification. that has been the biggest cost, because these meters are expensive, especially the big distribution meters.
So
Deep: How much of that is the number of, meters that you place and where you place them? Is that kind of just like engineering standard?
Ashwin: That's an engineering standard when it comes to this, water distribution network design,
Deep: Okay, and then as far as the smartness of the meter, is that largely like, hey, if you're buying a new meter, like, why would you buy one that somebody has to drive by to pull the data off? So it's a pretty straightforward decision to buy one that's connected to Wi Fi and get out the data.
Ashwin: The way we look at it, it should be a, a decision like that. It's like, okay, why would you pay somebody? Because the cost for the meter might be higher, but the cost for paying somebody to go collect that data over time is not worth that. over time the meter pays for itself. Right. So
Deep: largely you're, you're left justifying the aggregation and usage [00:34:00] and interpretation of that data and, operationalization of any insights you find.
That part you have to justify, but the rest of it sounds. Like a fairly straightforward, spend given that you're putting in a new water system. Is that correct?
Ashwin: That's fair. But again, we're also learning how to build us better. Right? So one of the biggest challenges people face in cities is, when trucks just back up to fire hydrants, steal water and maybe leave them running, right?
Hydrants are not monitored.
Deep: Oh, people still water like that. Wow. That's still
Ashwin: water. Yeah, it is. but it happens. we lost 3 million gallons of water over two days, two years ago. And that's one of the things that drove this. Okay, we need to build something because we don't. I mean, even 3 million gallons of water is few thousand dollars.
So it's not really like you're talking about millions of dollars, right? And in a city, that's not a big expense compared to what you're doing, but losing 3 million gallons in two days. If you just think about that number, that's [00:35:00] huge, right? it doesn't compute for me.
Deep: It makes me think of, I lived in Chicago.
I went to school there and in the summertime when it would get really hot. In the inner city, you know, it'd be like a hundred and something degrees and people would pop the fire hydrants, but you could have like a whole block where all the fire hydrants are going and then the kids are playing in them.
So what was the solution there to put sensors in the, in the hydrants?
Ashwin: So, yes, so that's what we're for the new hydrants that were, we would plant in the city. What are the, do we need a cap can alert when it's taken off? Do we actually need pressure sensing in the hydrant, or is it a mix of that?
Right? So that's the decision that that has to be justified. That's where the financial decision comes in. That's where the justification comes in. So we will do a mix. Some of those will have pressure sensors on some of the nodes and we'll work with the design engineer to figure it out.
Okay. These are the, the notes that we need pressure sensing in the hydrants and the rest of it on this loop can just have caps, that will alert when they're taken out. Right? [00:36:00] So those kind of design decisions would be, would be justified just based on cost. I mean, that's, that's
Deep: an interesting, that puts you in kind of an interesting position, right?
As a company, because. You de facto put out an RFP for, for a smart fire hydrant. So you might also be in a position to fund a company to build a smart fire hydrant, and maybe take some equity stake in that company.
Ashwin: We haven't even talked about this kind of thing. It's parking meters, right?
Parking is another big vertical. We're doing exactly that. What you just talked about. parking. I mean, I knew nothing about parking three years ago, but unfortunately, I know too much about parking today, and it's the same thing, it's a, it's a resource that's very difficult to build once you've already allocated.
It's like water. Once you've already built a plant to increase capacity, it's a fixed capacity resource. Parking is also like that. It's a fixed capacity resource. To increase it takes a lot of cost, right? So how do you build it right the first [00:37:00] time? Well, let's talk about
Deep: that because I feel like we haven't, we haven't really touched electricity at all.
But it seems to me, , if I imagine the future of parking from a city's vantage, it should have power ideally we have induction charging by then, but you've got e beads you're scraping 10 15 percent off that so that you, so somebody's making some money off the power that goes into the car.
and then on top of that, you need to know exactly where the cars are, which most parking systems don't, they're more rudimentary than that, you don't actually know which slots are full when, because you could ideally sell that data, to the car company, so you know exactly where to drive to get a parking spot or, and you could even imagine a reservation system kind of evolving on top of that also, that it's like integrated.
how are you guys thinking about it?
Ashwin: We see that problem from very different, Perspectives. Not singular, because we have the perspective of the people who are building parking. We have the perspective of the city who has to manage parking. We also live and work in this area, so we have the perspective of the [00:38:00] people who are going to be using the parking as well.
So we have to view it from all of these different areas, and the challenges are very different, and none of these three entities really like parking, right? Nobody likes it. from a real estate developer's perspective, that's land that is wasted. From a city's perspective, that is a constant source of complaints, and from a user's perspective, it's an annoyance at best, and a detriment to doing your stuff most of the time, right?
it's, it's really how do you manage it and make money off of it, like you're saying, right? Yeah, I mean,
Deep: ideally, if you're planning the city from scratch, you should be able to minimize car trips, you should have metro, you know, light rail, high speed chairlifts, like whatever, something, something else that's more efficient than having to lug your car around.
Ashwin: Yeah, so one of the things that we've done is that, like I talked about, this whole trail around the development, that's a walking trail with, you know, wider paths but being a realistic in Houston in the summer, nobody's going to walk, [00:39:00] right? And Houston's so much of a car centric city.
You cannot live without a car. You go everywhere. You're highly dependent on your car. if you look at the different levels of automation, you have, cars today and your parking systems have to support the cars today. But, you know, at some point in time.
You're going to have more self driving cars on the road than manual colors. How do you support parking? Nobody's going to put their hand out of the window to take a parking ticket. How do you manage parking in that situation? Right? How do you make your systems robust enough that you're supporting parking today and you're supporting parking
when you got other technology on the road so that that really plays into the systems that we put today, making sure that there's an upgrade path for technology going forward. When it goes to comes to parking and then also identifying. Okay. Like you said, which car is in which part like we build things based on a shared parking model [00:40:00] and try to peak shape, right?
If you look at power, peak shaving is a big concept and similarly for parking.
Deep: Do you mean reducing electricity demand based at peak times? Is that what you mean? Okay.
Ashwin: Yes, exactly. you look at, if you look at parking and you got an office building, like the building that I'm in, we have a parking garage that supports this building.
Now, that parking space is unused from 5 PM to morning, right? We have a hotel next door, so we build a hotel next door. We have allocated spaces and master plan so that , the parking that we build. Can be, the average use for that parking is over 24 hours, not at a certain point in time.
So we can build shared parking models based on land allocation and land planning and then also figure out ways to understand. Okay, we built this parking based on a certain model. Is it being used that way? Are we right in our model? How can we define our model? [00:41:00] And for that, we have to understand how our parking garage is used.
We have shared parking where certain people or certain organizations are supposed to park in certain floors. Do we know that that's being followed? And how do we follow that right now? So again, all of this is very interesting when it comes to sensors when it comes to how do we bring in that data and bring a holistic view of parking for the city rather than just okay, dislocation, dislocation, all that.
And now we haven't touched street parking again. Same thing.
Deep: I'm gonna change directions a little bit. So like on, your AI injection. We like to dig into basically three things. One is like what you're building. I feel like we got a good, sense of that at this point.
We've dug into there quite a bit. the second one is like how you're building it. We could probably dig in more there and maybe we will, touch into it, I think we have a reasonable sense of that. But the third thing that we haven't really touched on at all is, should we be building things this way? And this is like our entryway into, an array [00:42:00] of potential ethical, issues that arise. Let's fast forward 5 or 10 years out. You're successful. Let's say you have sensors in the water systems, the electrical systems , the parking systems, all the various aspects of the city. It's all kind of fed into dashboards that operators can look at.
So somebody administering the city wide parking system is able to sort of see real time views, heat maps. Bottlenecks, all that kind of stuff. Are you exposing their ability to, let's say, guide the city into, a positive direction?
And where are their potential, like, ethical constraints or conflicts that you either see or might see?
maybe Let's take a particular thing. Let's take consumption. Let's imagine you have limited water for a city. how does your solution enable the city administrators To achieve a goal of, let's say, 20 percent reduced water consumption.
Ashwin: For us, the reason that [00:43:00] we built out the service that I talked to you about. the driving force one is 100 percent accountability for watering and that's a goal.
That's I'm not saying that's practically achievable with all of the issues that we have in our systems, but that's the goal is to have 100 percent accountability for water that we produce, which means that the city knows We produced X amount of water from our plants.
We know where X amount of water went. Whether we build our residence for it or not is a totally different question. So first thing when we talk about resource and utilization for us, really, is understanding. Do we know how it is being used and where it is going? So first is understanding that. Second is minimizing loss. when we say accountability, we can still account for 100 percent of the water, but let's say 20 percent is a loss. Yeah, still not doing a good job, right? You can't count for it. But you know that 20 percent is lost. How [00:44:00] do you minimize that?
I think that is where a lot of cities are struggling with because of aging infrastructure, people, Not really having the tools and dials to understand how it's being used, where the losses are and how can you automate a lot of these, workflows to get these to minimize these losses. And that's where we're trying to provide those.
So let's say an anomaly happens. Today, it's a manual process. Two months ago, we had an anomaly detector in our non potable system. It took our operator nine days to fix it, even after knowing that there was an issue. We lost One million gallons of water in a field because of that.
It's a tiny need. We only lost one million, only one million, over nine days. If this was an automated process where, okay, we saw a leak, it was beyond a certain threshold, a service dispatch was automated because the location's right there, the person goes to that area, the truck goes there, [00:45:00] looks at the, finds the leak, fixes it, it's done in a day, or at best, two days.
now think about hundreds of such incidents in a city, right? Right. You're talking about effective, uh, effectively reducing loss and effectively reducing cost for water for the city.
Deep: yeah, and you could imagine a similar thought process applied to other resources it seems like an economist would say, well, just to have dynamic pricing on the water and your problem is solved, like every, all the actors themselves would self correct if the city can just toggle the, you know, like jack the price of water up in certain scenarios or certain times of year.
And then everyone's going to contract, assuming they know and they have visibility into it.
Ashwin: And assuming you have infinite resources, but that's the problem, right? Your water is not an infinite resource.
Deep: No, what I'm saying is going back to the hypothetical scenario.
I posed you as a city administrator need to reduce the water consumption by 20 percent in, over the summer. Cause there's a drought. taking it from [00:46:00] whatever, I don't know, a dollar for. I don't even know water prices, but like a dollar for a thousand gallons or something up to 1. 20. It seems like people won't even notice that.
In Seattle, I wouldn't notice that for three months because we get our bill every three months. So it seems like part of being a smart city would be, thinking through some of those scenarios and exposing that so that you could actually have consumption guided by maybe pricing or requests or whatever, and that you could monitor that in real time.
Ashwin: Yeah, I think the reason I mentioned the infinite resources, even in that scenario, even if you do that, the drawback that I see is that, it corrects usage, probably it brings down usage, but it does not bring down loss.
Deep: Correct.
Ashwin: It does not. It does not bring down loss. So that's why I was alluding to not being an infinite resource is okay.
You're still going to have all of that loss that you're you might be accounting for. You're not charging you're still wasting and you're still drawing whether it's from an aquifer, whether it's [00:47:00] from a groundwater resource, you're still drawing that water, right? And it's not Doing its intended purpose. And that really is what I'm getting at is, well, yeah, you probably do need dynamic pricing. Maybe not for, I don't know if you would do that for water. Maybe you would, but I mean, especially for parking and things like that. Of course you would. Yeah, right.
Deep: this has been a, a fantastic conversation.
I feel like I've, learned a bunch. I'm going to end on. my, kind of usual ending question, if we fast forward out, let's say 10 years, and your model is successful of a sort of preplanned smart city and all the sort of scenarios that you envision happen, What is the world look like or this?
Let's just say the city at that point. Like what? What does it look like? Maybe describe it for us, but also describe what it doesn't look like. what's the negative? The downside, you know, like, for me, when I think of what you're doing, I think of a town like Whistler, which was pre planned. I think Introwest was the developer, a lot of stuff [00:48:00] works, water works, electricity works, all that stuff, it's a beautiful setting, so nature's good, but it's kind of boring, all the buildings look the same, architecturally it's not interesting, it's certainly not like a Chamonix, France or something.
So it feels like there's there's inherent negative, as well that that come from such a perfectly planned city. So maybe describe your 5 to 10 year out vision, both the good and the bad,
Ashwin: our land planner would be livid. If you've suggested that to him, we have design guidelines for the city and it's very much, Not to be a boring city, right?
That that's really the vision for him being the master architect land planner for the city. But really, is a city if it's not for the people that are living there? I mean, it's it's really just a bunch of buildings and infrastructure, right? So for people to be here and wanting to be here, What people want really does not change.
What drives [00:49:00] people, what drives them to be happy, what drives them to want to be in a space is. Very primal, right? I mean, whatever makes humans happy. I don't think that has changed a lot or over time.
Deep: Well, I mean, I think modernization has played a role, right?
But like, generally, we want to be safe. We want to be able to afford, a nice space for our family. We want to be entertained. and we want to feel a sense of belonging and a sense of community. I mean, like those, those things seem reasonable.
Ashwin: Start with safe, right? everything works, it's a safe space without intruding on their privacy. That's that's one of the key goals for us is to be a safe space without including on privacy. And it's a very difficult problem to solve, though. And two, it's green.
If we have succeeded in our mission, then we're not dependent on your car to get around in the city. Whether that's through e bikes, whether that is through walking trails, whether that is through other micro mobility centers, [00:50:00] we've solved that problem if this is successful, right?
Whether that's through logistics hubs at the outside of the city.
Deep: Waymo's driving around, yeah.
Ashwin: Right. You know, things like that. We've solved the parking problem. We've solved the parking problem, not just for people driving, but for, you know, like you said, Waymo's driving around that, you know, we must are not driving around all day.
If there are more rainbows on the road than people driving, then obviously they're not driving around all the time. They need a place to regroup
we've solved that problem for mobility. we solve the problem for security. We've solved the problem for comfort indoors and outdoors, energy.
We've got energy goals. We've got water goals, all those. But if you look at it, really, the biggest per square footage cost in a building is productivity, the salary, how do you maximize that? And it's comfort. I think that's what I'm trying to bow down to is comfort.
Comfort is really key you know, we can have [00:51:00] all other goals, but if you've sacrificed comfort, people are not going to want to be in that building,
Deep: Well, it depends. It's like very few people are comfortable in New York, but they're all there anyway. And I think that's the part if I, if I had to like critique this approach, um, I think that the, the, the practical, efficient, all that stuff could be achieved.
But I think the vibrancy is a real question mark. I mean, no matter what you do, the fact that a city is built in the modern era only. means that it's not going to have that same soulful feel as a Granada in Spain or, you know, like it's, it's, it's, but that's okay. anyway, this has been an awesome conversation, Ashwin.
I've really enjoyed this conversation. Yeah, me too,
Ashwin: me too. Appreciate it. Thanks so
Deep: much for coming on the show.