Infrastructure Technology Podcast goes inside AT&T’s intelligent video technology

Key takeaways:

• AT&T’s internet of things video intelligence technology Is a game-changer: The technology dramatically reduces video data bandwidth—up to 10 times less than standard solutions and allows real-time video streaming over cellular networks, making it ideal for areas without fiber infrastructure.
• The deployment of the system is fast and reliable: Cameras can be set up in minutes, not weeks or months, and It supports integration with most existing camera hardware and video management systems, saving cities money and time.
• Stronger security and lower costs: Removes reliance on static IPs and VPNs, reducing cybersecurity risks and allows remote updates and diagnostics, cutting down on physical maintenance and truck rolls.
• Focus Is on infrastructure, not surveillance: The system does not have the ability to conduct facial recognition), and it’s designed to move video efficiently, not analyze identities.

In the season one finale of the Infrastructure Technology Podcast (ITP), hosts Gavin Jenkins, Brandon Lewis and Harlee Hewitt reflect on their inaugural season, highlighting standout episodes and their excitement for season two launching in September. Jenkins names his interview in this episode with AT&T’s Brad Miller.

Miller, who is the senior product manager at AT&T, discusses the company’s internet of things video intelligence technology. This solution uses an edge device and specialized software to stream video efficiently over cellular, satellite, Wi-Fi or wired networks and is particularly useful where fiber infrastructure is limited or nonexistent.

The technology aims to improve road safety, especially in rural and construction zones, by allowing cities and transportation agencies to monitor intersections and roadways in real time, improve incident response and even run analytics for future planning. Although not designed for facial recognition, the system can support analytics for detecting people or vehicles in restricted areas.

Here is a transcript from the episode:

GJ: And welcome to the Infrastructure Technology Podcast. I'm Gavin Jenkins, and with me as always, we have Brandon Lewis from Mass Transit magazine, Harlee Hewitt from Roads and Bridges. And ladies and gentlemen, we have a special, special podcast today because it is our season one finale, our 12th episode, end of season one. Brandon, give our listeners a little insight into that decision and tell them when we're going to be back.

BL: Yeah, so not only is today a special episode because it is Tuesday, it is a podcast day, we are back, but it is the season finale of the ITP, and this decision was made because folks, we just have had so many interesting and thoughtful topics, and we thought that we would come back strong in September with season two and give you even more great content than what we gave you in season one. But this season was filled with so many great memories, starting all the way back in January when we talked with Gary Smith about the supply chain. Then, we went all the way to rail infrastructure, then we went all the way to e-bikes. We talked about apps, we talked about mobility data, we talked about safety and security. Anything technology you can think of, we talked about, right here in the ITP

GJ: Harlee. I don't know how you can top that, but you know, tell us a little bit about your perspective from season one. Do you have a favorite interview, a favorite episode? What do you think?

HH: Absolutely. Well, first of all, Brandon, you really channeled a little bit of what we talked about with wrestling. That was intense. I know we spoke about that a few episodes ago, but as far as this first season, I mean, I think we built ourselves exactly what we wanted to. We built a great runway on our side, built up a really great portfolio for this first season of guests, and among them is our guest today, as we're going to talk about, was a great interview as well from Gavin, but I don't know. I just think that it was well-rounded, more well-rounded than I even expected it to be at first. We really covered a broad swath of topics, and I'm glad, and there's so much more to come, and I'm really excited for the future.

GJ: Definitely. Well, we're still getting used to the whole podcast thing, so we're still trying to smooth out the road, if you will. I'd like to just say that my favorite episodes all revolved around AI. I read news and reports about artificial intelligence on an almost daily basis. I am terrified of the future because not just, I think that it's a possibility that AI could wipe out the human race even before that happens. I think that I could be unemployed living under a bridge because of artificial intelligence. I think that it's an exciting future because it's reshaping everything, but especially our industries, and there's the possibility that it could make infrastructure so much better, and there's a utopia future down one road, and then there's the destruction and poverty down another road and hopefully we take the right path with AI. It's one of the things that we're going to cover on this podcast until we're replaced by humanoids, but I'd have to say that my favorite episode, my favorite interview that I've done so far this year is the one that we're about to air with AT&T. If I may dive into it a little bit here, I'm going to be interviewing Brad Miller, who's a senior product manager at AT&T, and he is in charge of the internet of things video intelligence. He has been working in this role for I think over seven years. He's worked in customer care, billing, operations, marketing. I mean he's helped really build this department up with AI. He's a very intelligent guy. This is a great season finale. This is like the season four ending of Seinfeld where they have the pilot, as AT&T is one of the core American corporations. When you think of American businesses, you think of and AT&T, and it's right up there. It's just iconic. And whenever I was a kid, listeners, I'm a little bit older than these two. They're in their twenties. I'm 44, so I grew up, I was born in 1980. I grew up in the eighties, and when you're growing up in the eighties, AT&T commercials would make you cry. They were just so emotional, so iconic. They showed how they connected the country. This was back whenever landlines ruled. When I was a kid, I had a rotary dial telephone in my house.

BL: My grandma had one. 

HH: Yep, same actually.

GJ: And during that time period, AT&T connected people, and now what they're doing is they're going to make intersections safer with video and cameras that don't glitch, and we're about to get into that, but before we get to my interview, any last thoughts before we dive in?

BL: Well, as always, I would just like to say thank you to At&T because if it wasn't for them, I would not be on here right now recording the ITP with my two wonderful co-hosts.

GJ: Yep, we got to thank AT&T and also want to thank you, the listener, for supporting us and listening to us through this first season, and we'll jump back on at the end and close out the episode, but first we're going to listen to my interview with Brad Miller from AT&T.

GJ: Bradley Miller, senior product manager of AT&T. Welcome to the Infrastructure Technology Podcast.

BM: Thank you very much.

GJ: Today, we're going to be talking about AT&T’s internet of things video intelligence technology, and it's being applied to transportation and infrastructure and what it contributes to safety and efficiency. So, with that being said, let's just dive right in. Tell me about AT&T’s internet of things video intelligent.

BM: Sure. So think of this as like a middleware solution. It utilizes a small edge hardware piece at the front edge right where your camera sits and then backend software that's delivering ultra efficient transmissions capable of streaming video over any AT&T network, including our LTE four, G 5G and Wi-Ffi, even wired at a significantly lower bandwidth than what is traditionally accustomed to being seen out there in the networks.

GJ: And I remember talking to you when we met. We met a couple months ago, and you told me that these are going to be placed in intersections.

BM: So the idea for transportation is it's ever growing, right?! The transportation keeps growing, keeps evolving and intersections are becoming smarter. Rural highways need visibility as traffic, and people continue to sprawl across the country into new areas that maybe or not quite have the infrastructure in place, so cellular really has kind of starting to become this go-to when the fiber isn't available or the existing networks that may be in place that are legacy networks just won't support the video at these fixed locations, so when you start putting cameras and hundreds of people at an intersection, and you've got cameras on top of that, and people are putting up one, four, six, eight cameras per intersection, the data network just is not really able to support video, plus everything else that all people and citizens are trying to do out there, so we really were trying to figure out how to develop a solution that is capable of being a good citizen, good citizen, right for the network, good citizen for our transportation customers, as well as good citizens for the citizen, right?! Because they need to be able to access the network when they come to an incident, right?! When something happens out there, everybody hits off of either a single cell site or maybe two, but when you've got all this video, it really constraints those networks to basically limit the amount of usable network for other customers, so the idea was to figure out a way to efficiently or optimize the video so that when it's transported, it uses significantly less bandwidth, therefore freeing up the network for everybody else to be able to utilize, very similar to our person at network.

GJ: Alright, so you've outlined some challenges there including data consumption, installation security risks. Can you tell us specifically those three? Can you tell us a little bit more about those challenges and how the informed IOT video intelligence will improve them?

BM: So when we look at the solution, whether it's from a cellular perspective, it's about ease of deployment today, and traditionally speaking, installation of fixed line video infrastructure is costly. Costly from lane cabling, carrying up the roads, concrete cutting, pavement cutting, but also trying to get fiber to a destination or to a single point. There may not be fiber close enough or a broadband connection close enough to do that, but there's more than that. It's not just about running the cable, it's the permitting and the time that it takes to do that could take months, could take years for certain construction projects. The construction itself and the personnel scheduling the delay, the traffic routing just continues to build a city's infrastructure budget for trying to complete a project. There's other things around challenges about temporary, again, kind of goes back to the same construction personnel scheduling, but even temporary construction sites are very safe from a perspective of the equipment's there but the workers that are present are not always in a safe environment so giving video and that ability to put video up within minutes versus days, weeks to plan out a whole project, they can now deploy this stuff so quickly but the other component you have in this is not just the cost of the infrastructure, it's the cost of using the infrastructure. So data usage is by itself a huge consumer of any network, whether it's cellular or not, but video can use 500 gigs per camera up to over terabytes of video data usage every month per camera. You think about putting city intersections will now have anywhere between four to six cameras per location. That's terabytes upon terabytes per intersection. How many intersections are in every city, whether that's an urban city or a suburban city? Downtown areas where there's an intersection every block, and you've got all these cameras now associated to that. It becomes a real data challenge to move that kind of data amongst everything else that's going on, so as they're adding more cameras, more lenses, higher end cameras, the bandwidth becomes a challenge. Then, you've got to secure all that video while typical public transportation or DOTS. DOT’s just stream video, right?! They're not recording it, securing that video, so someone doesn't tamper with the video or inject something into the video or the video camera creating malware, cyber attacks onto infrastructure that can, because video cameras are typically tied directly to a city's infrastructure, the edge device, the camera itself is by default a security risk though they spend a lot of time trying to make sure that the camera stays secure, the network is secure, their backend systems are secure, so you've got new VPNs, VPN licensing, IPSec tunnels, multiple network routing structures. You have to deal with static IPS mostly with video cameras, and static IPS by default are a security risk just because you can find them on the network so trying to get that out of the way, right, to be able to secure the video camera to prevent it from having malware injected into it or some sort of spyware or some sort of attack onto the camera, but also keeping it out of the customer's network from a standpoint of the security risk becomes part of a challenge. And then of course, you've got your ongoing cost, whether it's updates and maintenance, truck rolls, remote management, onsite management. Two people are required to typically get in a bucket truck and get up on top of to fix a camera that may be malfunction. What if you could remotely power cycle it? What if you could remotely update its firmware and do that through a secure connection without the need of VPNs and big static IPS all coming into your network? But these were all these challenges that the customers DOT’s and alike are all challenged with every day, and that's what the solution looks to help solve for different areas of their use cases.

GJ: Alright, so before we go on, I have got to ask this question. So, are these cameras facial recognition? Is our faces part of the data that's being collected?

BM: The solution itself is not about collecting data, it's about moving the video from point A to point B. The solution doesn't look to collect data. It's not collecting data. It doesn't collect biometrics. It doesn't collect anything about a person. It's literally the solution is about getting video from A to B as efficiently and optimized as it can, and then, where the video goes beyond that, the cities are responsible for that data. For me, facial wreck and doing personal object detection, that's not what I do. That's not what the solution is about, and it's not something that AT&T actively engages with from the Rema standpoint of looking at those types of characteristics or traits. 

GJ: The IOT video intelligence though uses Kodak in network aware technology that saves space without sacrificing video quality and that kind of technology can be really beneficial to preventing speeding and all sorts of other things when it comes to safety. How does that technology enable IOT video intelligence to work?

BM: So the Kodak, which is how all video is formatted, so video is transmitted when it's being in motion, it's transmitted or put into a Kodak for basically it's a compression algorithm, a Kodak is right?! So think of that like your traditional H264 video or now upcoming H265. H264 has been around for a decade or more. H 265 is the up and comer, and it's just a way to compress the video today as it gets transported. Where this Kodak is different is that the Kodak wasn't designed or developed off of the H264 standard, which is again what everybody uses. This Kodak was designed for cellular with this whole cellular concept as the premise. H264 was premised and designed typically on a fixed infrastructure, a broadband connection that has a very large capabilities, consistent, it's got low latency, it's got high bandwidth availability, and it's just ultimately constant and cellular networks, satellite networks are not. They're very constrained at times, dynamically shift throughout the time of day depending upon the event people, people attached to it devices, but this Kodak was specifically designed for cellular and satellite and because it was designed that way, it was inherently designed to adapt to the way cellular and satellite networks operate, which changes all the time, and it can change within just a few seconds, and so if you go from having five megabits per second or 10 megabits per second to only having 100 kilobits per second, a traditional video transport on H264 doesn't know what to do with it, so you get blur or you get chop. You get black screens, you get stutter. We'll just call it the rubber band effect, where it stops positive and then automatically speeds up, so that problem exists today, right?! So again, this Kodak that's part of the solution is designed to operate in those very dynamically shifting networks that have a constant change not only in the bandwidth, but by time of day of congestion, just number of users on it. So you've got this standard Kodak H264, and then you've got this patented Kodak, which is called TBI. That was designed for cellular satellite, making it hyper-efficient for networks and because it works on cellular and satellite and was designed for that, the same effect happens for customers who want to use this on a wired or Wi-Fi environment because its efficiency happens at the edge before the transport. The efficiency is across the board a trifecta in a way that it can work on Wi-Fi wired, but it's specifically, it's tailored at cellular. The other part of the technology, so the Kodak is the compression engine if you will, or efficiency part. It also uses artificial intelligence and machine learning algorithms inside of it so that it's actually doing two things at the exact same time. It's looking at the video that's coming in, optimizing that video based upon what's going on in the scene resolution range per second that it's being asked to provide, but it's also taking a backwards look at the network, so as that network changes, it can adapt the video so that it fits within the pipe or the bandwidth pipe that it's being given at that moment in time so that you don't drop the connection, it doesn't lose the quality, right?! It can shift and assess itself before it sends it so that it always gets from point A to point B reliably.

GJ: The video intelligence can integrate most existing management systems, software and IP cameras. How does that work and do the owners of the cameras, the cities, the townships, do they have to worry about that? The integration?

BM: So the solution itself as mentioned in thought is middleware, so it's designed to go between the customer's camera and that backend management system, also known as a VMS, or a video management system, so it's not meant to replace those two things, so we want the customer to use their existing infrastructure. It's a cost savings thing for them. The solution goes in the middle, therefore allowing their video to come through as it is today. They don't have to make any changes and the back in management system, their VMS, whatever that may be, it could be milestone, it could be a genetech, it could even be wow up or all like that's irrelevant to me. The system, the solution itself, is agnostic to those particular platforms, so when you integrate it, it's literally put the edge device at the edge with your camera, connect it to the backend server software that comes with this part of the solution and then that is connected to your VMS through call it a driver from their VMS over to this, so internally, it's once you have it in place, you can just continue to add edge devices at will and you just configure them. It takes five minutes, send 'em out the door. An installer doesn't have to have extreme knowledge of video systems. They just need to know how to plug something into power.

GJ: So once they're plugged in, and they're operating, how can a city or a DOT know if AT&T’s video intelligence will be an effective deployment tactic for that region, for that area, for the road?

BM: Typically for the customers that we've deployed thus far, we usually start with a small deployment, a lot of proof of concept. They want to test out the technology, they want to make sure that it works within their environment. They want to make sure that it works with, again, their management system, their IP based cameras, so we typically help them get started with that and the cool part about that proof of concept is that if we're going through this process of doing this as a small deployment, when it's successful, they don't have to do anything extra. It's already in place. They can now just drop and go, right?! They can just continue to scale, so we typically start with, most customers start between 10-20 units per. Small deployments may even see something as small as five, but again, once you have the infrastructure in place, which can be done in just a couple hours, maybe a day or two, depending upon how complex the customer's backend system is that we need to tie off into, they're done. They're ready to start testing or start evaluating the solution and generally what we've found is that the moment they see the video come up, they're like, ‘Wow, okay, that was way easier than I thought it was going to be, right”?! They didn't have to turn a lot of knob, they didn't have to configure all the cameras at once. I think just put it in and then we start showing them how they can reduce their bandwidth usage, and we usually do a comparison of what they had before, what they were using before and then show them here's what you're doing now, here's the usage you have now. You went from four megabits per second per camera to now you're only sending 250 kilobits per second, and they start seeing the math roll off in their head of, ‘Wow, I could do this for cellular now at 10 to one ratios from what they were doing before’, and then of course, the quality, the consistency in that video. They start seeing that they never lost the video feed, it never clipped, it never stuttered in a way that caused the rubber band effect. So really they start having this conversation of, ‘Well, okay, that's really cool. Can I apply this to X being wired’? Yes, we could do that. Just put the box on your wired infrastructure and it does the same magic. And then the next question is, ‘Well, we want to start running analytics now they have this visibility that it's consistent that analytics need to have’, so it starts to becoming a more solution, intricate part of their overall budget and overall conversation as they're planning expansions, not just both in the city, but on rural routes where we have customers that are using the solution for roadway stops emergency lane, whether it's a violation or they've got a vehicle broken down and being able to send response teams out there proactively versus waiting for the customer, but they also have visibility if it was maybe a dangerous scene to know what type of emergency personnel need to be responding to the event versus relying on 911 dispatch or a roadside assistance caller, right?! The other ones are they've deployed them in, as I mentioned earlier, like roadside, temporary roadside construction. They're using it for temporary construction zones to ensure the worker safety or in the event something happens, they being the customer, and DOT’s may have a record of that, of their own recordings so that they have what needs basically a post-analysis, a postmortem if you will, of what took place, how to prevent it going forward, right?! Workplace trainings. Some other ones are using it for evacuation route planning and giving increased visibility to areas or rural highways that had just never had visibility, so they didn't know if the traffic was stopped backed up, so there was an accident out there where emergency responders needed to be, and then you start walking off into the other customers that have started using the solution or using it on roadway signage, portable signage units where they're putting, again, kind of like temporary type stuff. Being able to rapidly deploy a video solution that is already integrated or tied into the central video management system is really catching their eye and attention because they can do this without having to go through long scheduling. They don't have to go order new equipment. They can just have this on hand, have it configured, hand it to the field installer, they go out, it's up as soon as they plug it in. It's up and giving them video in under two minutes.

GJ: You’ve mentioned rural areas a couple of times there. How does the technology help monitor even secluded areas? I mean, I know that the IIJA was meant to help expand broadband into even remote areas across the continental United States, but do the cameras have any issues when they're in remote areas?

BM: So that is where this solution kind of really starts to show its feature benefits or value to these customers is being in a rural area. You typically don't have the wired infrastructure that you need for high bandwidth need devices such as video and having it showing them that they can do this on cellular, not only on cellular, but do it at a bandwidth rate of under 300 kilobits per second, right?! You can stream the video using the solution as low as up to 10 kilobits per second, so when you think about rural, or you think about constrained networks, broadband being the constrained piece there, right?! Because as you get further out, maybe only dealing with copper or you're dealing with DSL types, needs video doesn't work like that, so this cellular, because cellular offers a greater bandwidth opportunity for these customers, it now gives them the option to say, ‘Well, I can just put it on cellular. I could put solar out there. I don't even have to run power. I could put this on solar, run it with cellular. I've got a completely contained unit and now I have visibility in an area that we've been completely blind to for the last five plus years’, so you could put it at a very rural intersection, the one stoplight count kind of thing, and you could now have video there where you didn't have video before, and then as soon as you have the video, then the next question is, ‘Well, what can I do to manage the traffic? How do I use video to analyze the traffic pattern? How many cars are turning left? How many cars are turning right? What's the lane usage? Are there people in the crosswalks running analytics or triggering smart traffic intersections?’ Being able to have that video constantly and consistently reliable plays a huge difference in whether analytics can truly benefit the end customer.

GJ: The technology can also help with safety to provide real-time situational awareness on roads and railways to improve incident response. Do you have any examples of success with that?

BM: So actually one of our customers mentioned just earlier, a little bit earlier, uses it for kind of their rural area. They had, I think it was like nine miles of road that was not visible from a video perspective. Again, no infrastructure in place there and what the public safety part of that is to be able to quickly identify when a vehicle is stopped on the shoulder. They can turn on the emergency signs to let people know that there's an accident ahead of, they can see it to alert tow trucks or their emergency roadside service vehicles to be dispatched to the closest location. Again, without the customer ever having to make that phone call, they're already enroute so reducing the time that a pedestrian may be on the side of the road, reducing the time that it takes for other drivers that may be coming up to a scene of an accident. Again, rural areas kind of being the play there, right?! You don't have video typically. The other part is, as previously mentioned, right, roadside construction for the construction workers themselves. Again, if you were running a video that you may put half mile up the road, and you see cars not moving over, or you see a car, maybe you're running analytics again, right?! You're getting constant video, so you could run analytics on speed and maybe it sends an alert to the construction site that a siren that goes off that just alerts them that there's a vehicle that's coming up on them that they need to pay attention to. One of my things that I feel is a little underserved is the railroads and the railroad crossings as we travel across America. Railroads are still the heart of transportation when it comes to moving goods, and you don't see a lot of video cameras on railroad crossings because there's typically no infrastructure. They still run on a copper contact that changes the track location from a switching control, but it's very, very rudimentary, and what if you could attach the camera to a crossing that could see a half mile down the road or right there on the crossing so that if a car stops for more than five seconds or 10 seconds, it alerts the railroad commission or the local city that there's a car stopped on the road? Maybe it's tied off into, again, the train, a locomotive, regional operator, whatnot. They have visibility. They can start slowing a train down long before the train sees the vehicle and prevent a catastrophe.

GJ: What if there is a mustache bandit, and he ties a damsel in distress to the tracks? Is there going to be a way to alert the engineer, the conductor on that train to stop?

BM: Sure. So again, video, we toss and turn around. Video is a sensor. Video is a sensor. It is really that, right?! Video is the sensor. The sensor is what you're trying to detect, so in the case of the pirate or the bandit mustache guy tying the damsel up on a railroad track, you could, because again, you've got video coming in, you could have an analytic on the back end that is used to identify a person, not who they are, but just the fact that they're human. They have two legs, two arms, they stand upright kind of thing, right?! It'd be the same applied to an animal. There are many analytics that can determine the difference between an animal and what type of animal versus a person, so it's doable, but yes, you could detect that a person or vehicle, right?! Both. It could be both in the same scene. Maybe they're going to run over them before the train gets there, I don't know, but yes, you could even do fall detection, so if you've got possibly an inebriated person traveling down the train tracks, and they fall over onto the train tracks, that happens, right?! You could have an analytic being run that said, ‘A person moved into this space, this is a restricted area.’ They didn't move for five seconds or 10 seconds, right?! Go ahead and alert the train operators on that line that there is a person on the tracks. They can go ahead and slow the trains down long before that incident actually happens.

GJ: What if you’re wrongly accused of a crime, and you're handcuffed, and you just want to put the chain on the tracks, you actually need the train to go over the handcuffs. Will it slow down or will it keep going and then, thus free you from the handcuffs?

BM: I would think that if you were running an analytic that was in play that would say you're a person of object size, and you're near the track, right?! Because you put what is called a geofence on the track itself, so that if anybody comes within X, and they cross a perimeter, that line, that geofence line, it automatically detects so the likelihood of you getting your chain cut, your handcuff cutoff, probably not. If you're running video and video analytics on top of it.

GJ: Technology as we know it is always rapidly advancing. Does AT&T have any plans to expand the internet of things video intelligent technology moving forward?

BM: Short answer is yes. As we've seen with this solution, there is a tremendous demand out there and a need to grow video. It's already growing on its own. We need to figure out the most efficient way to make it grow effectively while still maintaining the network performance that AT&T is looking for. Whether that's just straight up video on cellular using this type of solution, whether it's the employing other analytic solution providers tackle these challenges together, whether that's the Danville in distress, whether that's a roadside construction, whether it's train derailments or it's an innocent bystander on the road. All of these things come together, and they need connectivity to do it. We want to make sure that we have a solution and the technology combined that gives them the freedom to do and grow and scale, so again, we have today, right, cloud-based type video solutions from a future look kind of spark, and then of course, inclusive of analytics that a customer could look to deploy as part of their solution.

GJ: Well, Brad Miller from AT&T, the senior product manager, and this is such fascinating stuff, thank you so much for coming on and talking to us about it. I think that we have a lot of DOT and municipality people who listen to us and who read Roads and Bridges, and I think that this technology can really help out reduce crashes and reduce congestion, and most importantly, really for our industry make work zones safer, and I think that this is just an excellent piece of technology, and I appreciate you coming on and talking to us about it.

BM: You're very welcome. Thank you for having me.

GJ: And we're back. That was my interview with AT&T’s Brad Miller or Bradley Miller. If I was Brad, if I was named Brad, I would go by Bradley personally.

HH: Why is that?

GJ: I think it's because I was bullied in junior high by a guy named Brad.

HH: So you're holding the grudge. Okay.

GJ: I think the name. There are certain names that are just in my head. I just have bad thoughts of (laugh)… Alright, so that was my interview. Brandon, what did you think?

BL: Man, you know what, Gavin, again, as always, we've been here for 12 episodes, and as I tell you every time, fantastic Interview without fail, but this interview from the Mass Transit side, mostly from my perspective, Brad talked about that these cameras that AT&T is using that they're putting them in public transit agencies that these agencies use to stream, and they watch, and they look inside almost as security cameras, whether it's on rail cars or on buses, and the importance for those cameras to not only be accurate to record real on data, but to also not have any lag and for them to be secure, so that they don't get corrupted and things like that. But yeah, overall, again, great interview, Brad is very insightful and once again AT&T with that iconic globe sound effect whenever you hear that commercial, an iconic brand.

GJ: Harlee, before we jump to you, I just want to say that we're transcribing our discussion here in the chat and teams just thought that Brandon said his name's Brett, not Brad, and that is a really good name. I love the name Brett, but he goes by Brad, not Brett, but Brett. Oh my God. Brett. Now that is, that's a cool name. Harlee, this was another example of smart technology and how it can help us and make us safer. What were your takeaways?

HH: I think that the biggest takeaway for me was the reach, so he talked also about this being able to be installed a bit more easily than traditional systems in places like rural areas and rural highways, so I mean that's something that we've seen focus on in recent times. I remember that the Biden Administration released quite a few grants of the rural surface transportation program and most recently, that was in January, they were focusing on more of the infrastructure, roads and construction side. However, this is something that's very important. I mean, it's not exactly safe as someone who comes from areas like that to travel in some of those areas, so I think that this is huge because obviously AT&T most likely is in your hometown, so as you said, one of those iconic brands, it's out there, so they're able to put this wherever, and I think that's really going to help streamline some things, help even smaller transportation agencies and more with what they're able to do and in their processes.

GJ: Absolutely, and that's another big thing that we will be covering in season two is how our technology and the evolution of our technology is improving the lives of people in rural America and improving their infrastructure. That is definitely a bipartisan issue from our leaders in Washington, is they want to have rural America having better internet capabilities and better safety through technology like this, and so it's really important and thank you so much for listening. Harlee, thank you so much for all the work that you do on this podcast behind the scenes, and Brandon, thank you for all the work you do behind the scenes and also want to thank Ileana, who is rocking it, just absolutely crushing it as our new editor. Ileana is the Roads and Bridges digital editor and the producer of this podcast, and we also have to thank Endeavor Business Media. Endeavor Business Media is our parent company. They gave us the backing and the help in order to build the runway that helped us launch this 12-episode season. First season of the ITP and Endeavor Business Media is also allowing us to take a summer break and retool some things and come back strong and give you a season two, and we are so excited to do that. We'll be back in September right after Labor Day and any final thoughts to you, Brandon Lewis,

BL: If you have any suggestions on how we can improve season two, send it to us. Email us [email protected]. You can also follow us, Mass Transit, Roads and Bridges on all those social media channels. Facebook, Instagram, Twitter/X, whatever you want to call it nowadays. Contact us. We want to hear you over the summer because we are not going away. We're going to get back in the lab, and you're going to hear me, Gavin, Harlee and our friendly banter as soon as Labor Day, Tuesday, Sept. 2, 2025. So make sure you grab your leftover food, your hot dogs, your burgers, whatever you love on Labor Day and come sit down for a new episode of the ITP.

GJ: I completely forgot about social media, and yeah, the email, we have the email address. Please, please send us your thoughts, send us suggestions. Please email us. We love getting email. Thank you again to you, the listener for joining us on this journey, and we will be back in September. For Brandon Lewis and Harlee Hewitt and everybody else here at the ITP team. I'm Gavin Jenkins, and we will see you in September.