Smart City AI: Why Colorado Pilots Fail Before They Scale

Monisha Saldanha (00:02) Welcome to Colorado Tech People, the podcast where we explore the people, ideas, and technology shaping Colorado's future. Most people never think about infrastructure until something breaks, traffic congestion, unsafe roads, power outages, water shortages, delayed emergency response. But behind the scenes, cities are increasingly using AI, sensors, and connected systems to predict problems before they happen and operate more intelligently in real time. So what happens when cities themselves start becoming smarter? I'm your host, Monisha Saldanha. And today I'm pleased to be joined by Tyler Svitak Executive Director of the Colorado Smart Cities Alliance. In this conversation, we explore how technology is quietly transforming the systems that shape everyday life. from transportation and energy to public safety and urban planning. We'll talk about where smart city innovation is actually working, where it gets difficult, the trade-off cities face, and why Colorado is emerging as a leader in this space. Tyler, thank you so much for joining me today. Tyler Svitak (01:19) Thanks for having me. I'm excited for the conversation. Monisha Saldanha (01:21) Yeah, this is going to be great. So first question for you. Most people don't think about infrastructure until something breaks. What's a real world infrastructure problem people experience every day that technology is already helping solve? Tyler Svitak (01:38) Well, the first thing that comes to mind for me is potholes and just road conditions generally. I think it's one of the most notorious public complaints of their local elected officials is what the conditions of their roads are. And the way that it has been done since the beginning of roads is very different than how it's being done today using sensors and artificial intelligence to assess the real-time condition, and to continuously update that condition of infrastructure across an entire city at one time, as opposed to just scheduling when potholes get filled or responding to a call that is greatly skewed based on where they get calls from. Monisha Saldanha (02:27) And What are the mechanics for using AI to identify pot holes? How do you do that concretely? Tyler Svitak (02:38) There are many different approaches that companies are taking to doing this, but all of them pretty much involve a sensor. And it's usually a camera, but you could use LiDAR sensors as well. Some companies are doing that in order to take an image of a pothole or of cracking that's happening in asphalt. And... The most modern technology, even with cameras today, can measure the severity of that pothole, the depths of it, and ultimately give it a grade on what's called the Pavement Condition Index, which is a fun scale that road engineers use in order to assess the health of pavement. It's usually processed at the edge today so that you're not carrying all of that video and you're just taking a small snippet of that event that was triggered by the artificial intelligence and the computer vision that's happening on the edge of that device. And then the best systems are integrating it into asset management so that the maintenance team responsible for responding to a severe pothole can get that before someone calls it in and before it breaks your tire. Monisha Saldanha (03:56) Wow, that's really exciting. And I think all of us who drive on the roads will appreciate having the potholes solved before we ruin our wheels on them. Tyler Svitak (04:07) Yeah, I mean, it's one of the most prominent use cases where we're going from reactive to proactive. Monisha Saldanha (04:15) Fantastic. What surprised you most when you first started seeing how cities actually operate behind the scenes? Tyler Svitak (04:24) How complicated many cities are, and in fact, most cities or counties, most of us think about our jobs in terms of a business, right? And a business is very complicated. They have product teams where I know you grew up and specialize in design, you've got sales, you've got marketing, but in a city, you have potentially 50 to over a hundred different businesses operating under one organization. And that creates a lot of complexity in terms of systems and integration. And so if one business wants to do something, it needs to be able to integrate with the maintenance team, for example, that's going to go fill a pothole as opposed to the road building team that's going to build that road in the first place. so cities are really complicated and not because government employees aren't good at their jobs. These are incredibly dedicated people that care about their work, but they operate in very complicated systems. Monisha Saldanha (05:33) Yeah, that complexity, I can see how that would be a challenge. When people hear the term smart city, they often imagine futuristic technology. What does it actually mean in practice today? Tyler Svitak (05:48) There are literally thousands of definitions of what a smart city is that our listeners could probably Google. I think at the end of the day, the simplest definition that I can give it is that it's when government uses technology to change something for the better. There are always emerging technology tools I know that you cover on this podcast. you Government is not always the first to embrace those new technology tools. And so smart cities are ones that have designed the incentives and the processes for staff to take a problem like potholes, for example, and look at what new tools are available to them to try and approach that problem differently. And so it's not just cities, but it's counties. It's really smart government. And that's the world that I get to work in every day is with the innovation champions within those governments that have a problem and are looking for new ways to solve it. Monisha Saldanha (06:51) What's one example where technology is already improving daily life in a way most residents probably don't even notice? I know you already gave the example of the potholes, but is there another example that you could give? Tyler Svitak (07:05) Absolutely. Yeah, something that's not necessarily cutting edge and new, but is very helpful to all of us in times of need is when an emergency vehicle goes through a traffic intersection, it has what's called emergency preemption over that traffic signal. It's a wireless communication between that emergency vehicle and the traffic signal that turns all of those lights red. More recently, things like snowplows and transit buses use what are called signal priority, where it doesn't turn all of the lights red except for the one that's going through. It actually knows when that vehicle is going to approach the intersection and advances the green light for them so that they continue their journey more quickly as opposed to canceling the whole intersection, which is what pre-entry does. And so that's very commonplace. across most jurisdictions at this point. And it's something that saves people's lives, literally. Monisha Saldanha (08:10) Here's a question for you. Where I live, I have a traffic light where there's like a main road and then there's like a side road that comes in. Are there likely sensors that are monitoring whether there's traffic on the side road? Because I always wonder that because it's, you you don't want to stop the main road if nobody's actually waiting on the side road to get in. Tyler Svitak (08:36) Great question. Yeah, our transportation systems are complex and the answer to that question depends on probably who owns that road and the technology that those traffic signals are using to manage the road. So yes, the answer to your question is yes. There are sensors on most urban roads that are determining traffic flow and traffic time to get through a corridor, but it doesn't always necessarily inform how that light turns red or green. Now, the cutting edge of what transportation systems are looking at today is taking detection technology sensors on the entire roadway system. So your arterial road that feeds a main highway, for example, but across, let's say, 100 intersections. And in that case, artificial intelligence can take that data and optimize in real time the best traffic flow for the system and change the traffic signal timing accordingly. That's not the way most systems work today. They work on set schedules, no matter what the real time conditions of the roadway are, but we are very much moving in that direction. And, you know, there are essentially digital twins for transportation systems that are rolling out in cities today. Monisha Saldanha (10:01) Wow, that's super exciting. Tyler Svitak (10:04) It is. Monisha Saldanha (10:05) Looking ahead five to ten years, what infrastructure problem could feel dramatically different because of AI or connected systems? Tyler Svitak (10:17) So many things. think Infrastructure, when I think about this question, is both physical and digital. So whenever you go to apply for a permit to build a home, for example, or to build a new community that will become part of the housing infrastructure of a city, that whole process today is quite manual and human. And so if you're getting a permit for anything in a city, whether it's to build a house or to dig up something in utility right away, that process is very much going to be different five years from now because artificial intelligence is dramatically increasing the pace at which those applications can be reviewed, compared against regulations, and then approved or denied. And so there are many communities in Colorado, Denver, Mesa County, Colorado Springs and Centennial are all looking at artificial intelligence type systems that can help assist a human more quickly review permits for infrastructure changes. Monisha Saldanha (11:31) It's amazing to think of the impact of AI in the coming years. At a practical level, Tyler Svitak (11:38) It is. Monisha Saldanha (11:39) what technologies are having the biggest impact today? AI, sensors, IoT, real-time data systems, or something else? Tyler Svitak (11:50) I think I don't want to be the squeaky wheel and say AI, but at the end of the day, I think what artificial intelligence does that those other technologies don't is it maximizes the value of all of those other individual technology pieces in the stack. Smart cities as a brand really came out of the internet of things, right? And you know, that 2010s era type of technology that allowed devices to communicate with each other and to control things remotely. But without artificial intelligence, what a lot of cities found is that it created a lot of data, but it didn't give them a lot of insight or information from that data and certainly didn't allow them to use and act on it in real time. What artificial intelligence does is of course, it allows you to take all of that information to gain insights to tag specific events that require human action and intervention. And so I think artificial intelligence has unleashed the full value of all of these other technologies when it's integrated into that stack. Monisha Saldanha (12:56) And what about the costs associated with implementing these technologies? How are cities which are operating under budget constraints able to afford to implement these new technologies? Tyler Svitak (13:11) Well, oftentimes they aren't. It's definitely one of the biggest barriers facing the industry is, you know, governments don't have large capital markets that they can go to in order to take risk and take bets. In fact, many communities and constituents prefer that governments don't take much risk. so budget is a huge issue, but a lot of the innovation that happens in the smart city space usually begins with grant funding. So there's, you know, both local, state and federal grants that are available for technology. In fact, the transportation reauthorization bill that I believe just passed yesterday, US Congress is going to for the first time allow for federal transportation dollars to be used on what's called digital infrastructure and technology that includes things like artificial intelligence. Funding is always a concern. And one of the things I do is help communities identify where funding could come from in order to support the first trial of something. And then you got to prove it works and have some outcomes that you then go and compete against the other 50 business units in a city for some general fund dollars. And that's always the hardest thing is taking a pilot and scaling it. Monisha Saldanha (14:36) And Taking the example of potholes, how do you prove that the money invested in identifying potholes predictively is worthwhile? Tyler Svitak (14:52) Great question. Ultimately, potholes have a line item budget, right? So does road maintenance generally and road paving. So in this use case, I think, you know, the outcomes are actually relatively easy to define and determine if you can measure them over multiple years. So for example, can I fill more serious potholes with the budget that I have this year, and can I measure a decrease in the number of calls and the number of complaints that I receive from constituents? Because in government rate, it's not always about money. It's not even always about time savings. It can also be about constituent satisfaction. And so if you can get measurable improvement in both or all three, then that's a really valuable ROI for a government type of innovation project. Monisha Saldanha (15:53) If somebody damages their car on a pothole, is the city held responsible for that? Tyler Svitak (16:00) Not that I know of. I'm not aware of any case law that someone has sued a city for a pothole in the past, but potentially in the future, right? If someone could argue that you should have known that pothole existed and there's technology for that. And I called and complained about it. Who knows? Maybe that is a future liability issue for communities if technology advances. Monisha Saldanha (16:28) Interesting. How are cities moving from reactive systems to predictive systems? Is this a trend? Tyler Svitak (16:38) It is an ambition. Absolutely. I think most of the communities that we work with cite going from reactive to proactive as one of the incentives for, you know, dipping their toe in the smart city space. What, you know, IOT and artificial intelligence and digital twins, all of these types of technology tools allow is to not be limited by the systems or the human capacity of what is currently in place today. Of course, you your public works team of whatever, 20 people, isn't gonna know where every single pothole is. It's just not possible for them without the help of technology. And so that is one example of how a community can go from reactive to proactive is receiving a call from someone that's angry that there's a pothole, that pothole already exists. But in the case of, pavement condition index monitoring, if I know what the condition of every inch of asphalt is in my city over time and at all times, then hopefully I can work through my backlog of potholes and then figure out how to better maintain and when to pave my roads so that I get fewer and fewer potholes over time. And eventually, hopefully it's not an issue. There's of course budget and other considerations there, but in theory, that is one example where you could be very proactive. Same with water infrastructure. Water main breaks are a huge cost for cities and sometimes the power has to go out, water systems get shut off. It's a big expense, it's a big pain for people. The same type of technology that is assessing potholes and road condition is also assessing underground water infrastructure so that you can more proactively maintain that and avoid a really costly main break. Monisha Saldanha (18:32) With the water example, is that primarily LiDAR technology? Tyler Svitak (18:39) You could do it with LIDAR. The companies that I'm familiar with that are doing it here in Colorado are using cameras and artificial intelligence because that is how water infrastructure is maintained today. Actually, CCTV cameras are put underground. And then you have water engineers that look at that video, the inside of a dark water pipe, and determine and classify defects. AI is of course great at identifying defects and events. And so, you know, these engineers can now focus on fixing those problems and designing the right fix as opposed to having to spend their time sitting in a computer looking at the inside of a water pipe. Monisha Saldanha (19:25) Really interesting, and are those cameras fixed or are mobile inside the water pipes? Tyler Svitak (19:32) It depends. There's different ones. In fact, there are drone mounted cameras. So these are protected drones that can fly on the inside of a water pipe. And they are inherently detached, untethered. And then of course, there are tethered ones. And then the ones I'm most familiar with are just hardwired cameras that can reach a certain distance into the pipe. And then they have to go to the next manhole. Monisha Saldanha (20:01) Really interesting, I'm learning a lot in this conversation. What do people misunderstand most about smart city technology? Tyler Svitak (20:13) Ooh, there's a lot of misunderstandings. think A lot of people are afraid of government using technology for issues of privacy and enforcement. And I think that there's a big misunderstanding there that government, especially local government, city government, has any motivation, incentive, or time in the day to be proactively looking for you know, for surveilling their constituents. It's just not the way it works. And so that's one big misunderstanding. I think the other big one is that it's not really about technology. You know, I think a lot of critics of the smart cities movement feel like it was the industry that created this momentum for technology and government. And that started us on the wrong foot. In reality, these are just tools available to solve problems that matter to you and I every day. So Smart Cities is not really about technology. It's about people. It's about the systems in government that technology can improve. And technology is just kind of a necessary component of the larger ecosystem that cities operate within. And I think that's a big misconception that smart cities is all about technology. It's not, it's about solving problems and changing government. Monisha Saldanha (21:50) And what sounds exciting in theory about smart infrastructure, but becomes much harder in reality? Tyler Svitak (22:01) Yeah, I've been involved with well over a hundred different smart cities projects testing all sorts of solutions to all sorts of problems and public safety and transportation and water and energy, and parks and recreation, right? There there's opportunities for technology to improve any type of system and any type of government. What is much more complicated and as you as a product you know, owner and developer understands that the way that government functions today is that they pretend they know exactly what solution they need and they go out and buy it through what's called a competitive request for proposals process. And that requires them to know what they're buying. In the case of technology, they may not understand how artificial intelligence works. So pretending to go out and buy something that they don't understand or don't even know. aware of is it's just the wrong way to go about things. Businesses are developing solutions that they don't really understand the problem for. And that's because there's no real bridge for government to collaborate and do any user research, right? For example, on what that product needs to do and how it needs to integrate with other systems. And so what I see a lot is, you know, a point solution to something that doesn't scale, cause it doesn't integrate with the other solutions that are currently in place. The asset management system that the pothole AI needs to integrate with, for example. And so cities are complex, the systems they operate within are complex, and that complexity doesn't usually have the time to be considered on the front end of product development in today's environment. And it's one of the things I'm trying to help change. Monisha Saldanha (23:58) Well, that sounds like a big project. Tyler Svitak (24:03) Yeah, I mean, getting government to be vulnerable and to collaborate with industry and trust that they're going to build the right solution, you know, it's hard, but it's a lot of fun. And there are a lot of good people out there trying to build the right solution to improve communities for sure. Monisha Saldanha (24:24) We talked a little bit earlier about the limited budgets that cities have. They also have competing priorities. How do leaders decide which problems technology should solve first? Tyler Svitak (24:38) Ooh. I mean, that's the meat of smart cities really is. what direction do you point innovation toward? What problems are important enough to solve and to take risk in solving? And every government is and probably should be taking a slightly different approach to prioritizing which problems they want to solve first. We work with about 40 different government agencies across Colorado, and each of them have different priorities. One community might be very focused on public safety. Greenwood Village, for example. Another community might be really focused on transportation. Denver has a big focus on transportation, both safety, ensuring people aren't getting killed on our roads, trying to walk to school or to the store, or just trying to get to work in their car. So every community has different priorities. And that is the magic of smart cities is that every community is different. and you can apply technology to any of those problems. So we help, we try to help them prioritize which problems to focus on and to start with based on which ones are most important and have the biggest risk of keeping the status quo. Monisha Saldanha (26:03) It sounds like walking a tightrope. Tyler Svitak (26:08) A little bit, yeah, but you got to start somewhere. And I think taking risk and trying something new should be prioritized for the problems where the risk is already really big and the risk of failing is lower in that case. Monisha Saldanha (26:29) Can you share a Smart City initiative that didn't go as planned and what that taught you? Tyler Svitak (26:37) Absolutely. We have had many projects, I would say of the 110 or so that we've been able to facilitate at the Alliance. And there's probably twice that many that started and failed for one reason or another. One thing that I know we learned a lot from I don't know if I would consider it a failure or not, but I'm sure a lot of people would. It was a very media-worthy project where we put seven fully automated shuttles onto public roads at the Colorado School of Mines about five years ago in a service called the Mines Rover, where there were electric, fully automated shuttles transferring students and people between downtown Golden and the School of Mines and some student housing. Excuse me, and these were slow moving shuttles. And they didn't necessarily navigate the environment perfectly all the time. And it caused a lot of consternation with the driving, the motoring public, because they had to sometimes get stuck behind one of these shuttles that don't drive as fast as they do. And sometimes they would stop in the middle of the road if they couldn't navigate the environment confidently. That mixed with some maintenance issues that we had that didn't keep the shuttles as frequent as we wanted. Ultimately, that project shuttered earlier than it was supposed to. It operated for six months instead of the full year, and the school decided to shut that project down. But we learned a ton about where the market was at that point in time and what this tool could be applied to, the types of problems that it could and cannot be applied to successfully. And so in that way, I think it was a success for the industry for sure. Monisha Saldanha (28:40) That sounds like some important lessons learned. And One of the hardest things in innovation is scaling beyond pilots. And you mentioned that earlier in our conversation. Why do some smart city projects struggle to scale? Tyler Svitak (29:02) Well, because when something scales in city budgets, something else has to diminish, right? Unless, especially in Colorado, where we have the Taxpayer Bill of Rights and our tax base doesn't go up very quickly, there is generally a constant budget or a budget deficit. And so you have to convince either your department head or other departments that that innovation, that pilot you tried, is gonna be so effective that it's worth cutting something else. And that's a hard argument to make, especially if during the pilot phase, either you didn't design it so that you knew what outcomes you needed to make that argument, or it didn't run long enough, or there's a lot of reasons why you wouldn't have the data you need to go and convince your peers that you know, the mayor's office that this is worth cutting something else. And so that's the biggest challenge. And then there's also a lot around procurement that we probably don't have a ton of time to get into where something is piloted under a certain procurement process. And then scaling that process might need to be a different procurement, you know, altogether. And so procurement is something that's really sticky in government that industry doesn't have to deal with as much. Monisha Saldanha (30:37) Colorado is becoming a leader in smart infrastructure innovation. What's driving that momentum? Tyler Svitak (30:47) Well, I think Colorado, because of things like the Taxpayer Bill of Rights and the fact that our state doesn't necessarily have as many resources as other government agencies and other places, it's forcing, it has forced our governments to find innovative, creative efficiencies that other states may not have to. Also, I think we have done a great job of through the Alliance that was the first state nonprofit created to build this type of government innovation ecosystem in the United States. Just the existence of this ecosystem has allowed one government to try something and share it with another. No one city can try everything. And so by creating a community where peers are doing things and learning from each other, we've been able to replicate what works and share successes across. And so I think our brand is stronger. It's greater than the sum of its parts because we've been able to unite it a little better than other regions in the country. Monisha Saldanha (31:58) And How do smaller or rural communities fit into this future, not just major urban centers? Tyler Svitak (32:07) You know, my meeting just before this was a peer group we have on small town innovation. And, you know, there are actually a lot of advantages in small and rural communities for innovation to take place that big communities struggle with. For example, they don't necessarily have as much process, as many systems, or as many politics to navigate in order to get things done. And so one of the things that was uh, said today by um, someone, an assistant city manager. It is like, if you have an idea, my city manager says, hold up a mirror because you're the one to do it. And so people are very empowered in smaller towns to accomplish things. And if you have a new idea, it's easier to get done. However, in a small town, you don't have as much human capacity, you may have even smaller budgets. And so you have fewer resources to do those things, but fewer constraints as well. And so I see a lot of opportunity for smaller towns to innovate. One of the most notorious ones is a place called Peachtree Corners in Georgia. They operate what's called the Curiosity Lab. And this is essentially a playground for transportation technology companies to come and test their technologies on a city infrastructure. So if you wanted to test a camera and how it detects crashes, for example, you'd need to hang it on a traffic signal and that's hard to find. And so they have specialized in innovation for the public sector by providing their infrastructure to companies to come and do testing and the innovations that work they pick and choose from. So it's a small community and they've built a global brand for smart cities. Monisha Saldanha (34:02) What excites you most about the future of infrastructure over the next decade? Tyler Svitak (34:10) That it's changing, that there are now tools available to infrastructure owners that did not exist before to help improve their jobs day to day. And with artificial intelligence, they don't all necessarily cost an exorbitant amount either. There are a lot of solutions using AI that can use your existing hardware, your existing infrastructure, your existing cameras, for example, on a traffic signal, and provide intelligence and information to you without an enormous capital investment in new technology. And so I'm very excited for that. mean, Things like real-time transportation systems that can change traffic signal timing and adapt to the real-time conditions. And we're doing a lot in the emergency management space as well. Think about a fire, for example, especially in Colorado on our foothills. If a fire is sparked in one area, today there are sensors that can detect that fire instantly that we're deploying in Colorado Springs, for example. And then AI can model how that fire is going to spread based on the real-time conditions on the ground. and proactively evacuate communities before and with as much notice as possible based on where that fire is going to go. And so there's a lot of technologies that will save lives and also hopefully save, you know, car tires as well. Monisha Saldanha (35:57) What's one thing people should start paying more attention to as cities become increasingly connected and data-driven? Tyler Svitak (36:09) I think governance is incredibly important within governments and is varied across them. One very prime example in the media today is a company called Flock that uses automated license plate reading technology in order to flag events and where people are based on an event to law enforcement. That has become a hot topic here in Colorado because one of the owners of that data, or not owner, but excuse me, consumer of that data is the federal government. In the case of Flock, they have the opportunity to share that license plate reading data with ICE and with Immigrations and Customs Enforcement, and it's created a lot of issues. So the governance of what data is collected, who has access to it, how it can be used, is gonna be very, very important within cities because the Public Works and Transportation Department has a very different mission than law enforcement does. So determining who has access to the pothole camera data, for example, is gonna be really, really important because there is a huge difference in perception versus reality of what the public's gonna believe cities are using this technology for if it's not written and clearly communicated and published of what the governance of these systems is going to be. Monisha Saldanha (37:52) Governance. Yeah, very important topic. What's one book every builder should read and why? Tyler Svitak (38:04) This is a tough one. I have two. The first one though, I think everyone can relate to is called Killed by a Traffic Engineer. It's a book written by Professor Wesley Marshall of CU Denver. It's ultimately a really plain language read about how the transportation system we all navigate was designed by traffic engineers that were trained on not a whole lot of science. And so the reason that so many people die on our roads, over 40,000 people a year, which is more than any other country by a long shot, has as much to do with how our roads are designed as it does with how we drive on them. And in fact, how we drive on them is greatly influenced by how the roads are designed. And so it's a great book, very well written, hilarious, and I'd recommend anyone. Anyone read it. Monisha Saldanha (39:07) And what's your second book? You can have two. Tyler Svitak (39:10) Okay. The second one is called Recoding America. It's written by Jennifer Polka, who was the founder of Code for America and one of the leaders of the US Digital Service, which has embedded technologists inside of the federal government to help with things like the healthcare.gov rollout back when the Affordable Care Act was passed. And so it's all about how government particularly the federal government, should be better at technology and become more of a product owner, for example, as opposed to a project owner. And it's a great read. I think there are multiple politicians on the cover of the book talking about how every elected official definitely needs to read the book. But I think it's very much relevant to just anyone in the United States. Monisha Saldanha (40:07) fantastic. Well, those are two new books for me, so thank you for adding to my list. I look forward to reading them. Tyler Svitak (40:15) Of course. Monisha Saldanha (40:17) And thank you as well for joining me on the podcast today. It was a wonderful conversation. Tyler Svitak (40:23) Yeah, thank you for highlighting what I think is a really important topic that affects everybody, whether they know it or not. Monisha Saldanha (40:30) And thanks to our audience for listening to this episode of Colorado Tech People. Most of us rarely think about infrastructure until something fails. But as cities become more connected, predictive, and intelligent, the system shaping everyday life may begin changing in ways we barely notice until they fundamentally improve how we move, live, and interact with our communities. If you enjoyed this conversation, share it with someone interested in the future of technology, cities, and innovation. And don't forget to subscribe for more conversations with the People Shaping Colorado's future. Until next time, keep exploring how technology is changing the world around us, often in ways we cannot see.