Scott Stanford:

Hey, hey. What's up everybody? Welcome to The Infrastructors, the premier podcast for engaging conversations with influential thought leaders in AI, tech, government policy, and smart city innovation. Today's guest for Safe Fleet, Justin Malcolm. Justin, welcome to The Infrastructors, my friend. Thanks for taking a few minutes. Appreciate the time.

Justin Malcolm:

Thanks for having me.

Scott Stanford:

Well, tell us, you know, tell us about Safe Fleet. I mean, the mission for the organization, your technological portfolio really is just, you read about you guys, it's just transforming these fleet vehicles across the globe.

Justin Malcolm:

Well, so, you know I think sometimes I find it's hard to talk about. At the same time, the mission's so easy 'cause we're so involved in so many different things, as you said. So, you know, all kinds of specialty fleets. You mentioned a lot of them in school buses, transit trains, you know, fire, emergency, police you know, waste, recycling, commercial vehicles, you know all kinds of stuff. So, but, and everything from electromechanical, lighting, doors, hatches, you know mirrors to, you know, very advanced sensors and lidars and radars and cloud technology and all that kind of stuff. But, really our mission is always the same. So our, you know, our goal is to make, you know, these vehicles safer and in and around the vehicles to be as safe as possible to make them as productive as possible. You know, and I think that we say that, you know, "We try and make smarter vehicles which lead to better drivers, which lead to safer fleets and that that leads to safer streets." And that's the mission. So, you know, when it comes right down to it, that's what it's all about.

Scott Stanford:

What's been the biggest advancement in that area of your business?

Justin Malcolm:

Well, so I think that, you know, is really adding sensors around the vehicle, making the vehicles more intelligent, really giving information to the drivers about what's going on around them. You know, these large vehicles, you know, they can be, you know, 40, 60 feet long. They have a lot of blind spots. There's just lots of places the driver can't see. And so, providing systems that give you information about what's going on in the areas you can't see and making sure that you don't do anything that's unsafe. I think that's been probably the biggest advancement. And that's the area that probably continues to advance the most. We've been involved in video for a long time, so we were, you know, originally putting like VCR tapes, if you can imagine VHS tapes on buses. But yeah, it's, you know and then video has been used, mostly passively in the past. And so the real change is now that we can use a video more like a sensor and do active analytics, actually analyzing what's in the video. You know, because before unless a human being looked at it, you know, then nothing was captured even if there was an unsafe event. Whereas now you can be constantly monitoring that. You can be taking action if something happens. You can be, you know, creating statistics about what's going on. So, you know, that's really becoming, you know more and more sophisticated as time goes on.

Scott Stanford:

Yeah. And, you know, we were talking about it earlier before we even started how, you know, we're kidding around about how, you know, AI, artificial intelligence is really, really creeping in, and it's gonna be even more prevalent in our daily life. AI based technologies, are there a lot of those that you guys take advantage of right now for you know, at Safe Fleet, you know, where you're just achieving your mission? Or it's you're slowly moving into it, or is it really something that you guys haven't gotten involved with just yet?

Justin Malcolm:

Well, no, so it's actually, so we definitely have. And, you know, it's kind of both all over the place already, and then feel like it's still very early days. So, you know, it's evolving really quickly. But yeah, we have AI all over everything we do. So you might be, that might be driver assistance systems, maybe simple systems like collision warning or as I mentioned earlier, maybe blind spot elimination telling you there's pedestrian at risk somewhere that you can't see. We might be collecting data that's actually flagging events which are unsafe which can become a coaching opportunity later. You know, review of maybe that data after the fact to see if there's maybe the route of a fixed route vehicle has to be changed. Where it stops is dangerous. If you could eliminate a left turn, maybe that makes the whole route safer. You know, we might be warning children when it's unsafe to cross the road, to get on and off a school bus, or warning an officer that it's unsafe to approach a stopped vehicle alone. You know, there's, we might be looking at issuing tickets automatically in order to, you know, keep buses and bike lanes clear, you know, which has an impact on an entire city. So, you know, there's lots of places where we're using it today but at the same time, you know there's just an endless, you know, list of opportunities. And as you capture more and more data for more and more sources.

Scott Stanford:

Yeah.

Justin Malcolm:

But as AI gets, moves beyond just being able to find patterns to do things that are more creative, you know, there's, you know, a lot of this data that we just, otherwise just impossible for us poor humans to kind of wade through. AI can, you know help us summarize that data, pull out what's most important of it, actually make recommendations. And so that's, you know, we're gonna see more and more of that kind of stuff happening as well. So, it's just, there's a lot going on.

Scott Stanford:

Is there one particular technology that you're hearing about that's coming down the pike from the engineers that you guys are excited about for the future? Is there something that you guys are really looking forward to that hasn't happened yet that you know of?

Justin Malcolm:

Lately, we've all been hearing more and more about AI, but it's really, you know, what happens almost at scale. Like I said, as you get more data, as you get more processing power, it's like it's something magic happens where it's, you know, turns into something more than the constituent parts. And I think that that's kind of the exciting thing is really how do we pull all these things together? I mean, even when you're thinking about, you know on the road or at the edge, it's just more effective when you pull in things from different systems and combine those together, because you can, you know more about it than you can from every one of them. You think about the way that we work, you know, we use our five senses to figure out what's going on in the world. You know, so if you take away any one of them, if, you know, if I can't see, or if I can't hear, then, you know, my ability to react to the world to stay safe, to make good decisions is, you know, greatly impaired. And I think that, you know, we're seeing the same thing. And what we realize is right now, we've got, you know, systems and vehicles and drivers that are, you know running around. Like they've got, you know, only one sense, and you know, we need to give them the five.

Scott Stanford:

You know, Rekor Scout is crucial in Safe Fleet's able solution for realtime ID of vehicles. Can you give us a few examples of real world applications? And maybe, and as you were just talking about those senses, how it does change driver behavior?

Justin Malcolm:

Sure. So yeah. So you mentioned what a really great example actually of where we're using the Rekor technology is in this ABLE system you mentioned. So, ABLE stands for Automatic Bus Lane Enforcement. So you imagine in major cities if you go operating a transit system, traffic and congestion is a major problem, right? So you can't get your, you know, bus to get where it needs to go if the lanes are blocked. So a common solution is to create a dedicated bus lane. But, what happens is that if people are parked illegally in those bus lanes, then actually they become less safe and actually the buses operate even more slowly than if those lanes weren't there. So as they have to navigate in and out of other congested lanes, there's opportunities for collisions or, you know, incidents with pedestrians. That obviously takes a long time. So bus lane enforcement is a really hot topic. How to make sure that you keep those lanes clear. But you can imagine, you know one of the cities that we operate in, and it seems-

Scott Stanford:

Sure.

Justin Malcolm:

Almost impossible to really enforce all that. So, we have a system that's uses cameras and AI based, you know, powerful AI computer on a bus and some communication equipment that allows you to connect that vehicle in real time to the cloud. And it works automatically to detect when vehicles are parked in bus lanes and to issue citations. So the number of citations you can issue, the consistency with which you can enforce those rules definitely changes driver behavior. So, I can go into how that system works if you want, but to your question about how it changes behavior, you know, probably the right way of looking at it I think is even, I mean, it's obvious that if people are getting ticketed, and the data definitely shows that they're less likely to, you know, to break those rules in the future and to block the bus lanes in the future. But, where you really see that, I think, is in the impact that it has on the system and on the city. In lanes where this automatic bus lane detection and ticketing technology has been implemented.

Scott Stanford:

Sure.

Justin Malcolm:

We've got a product Clear Lane. You know, what you see is that the buses that operate on those routes are involved in fewer accidents even though they travel about 30% faster. And what's really interesting actually is 'cause after a year, if you have a route which is featuring this technology, you actually get about a 20% increase in ridership, which makes sense. I mean, you're gonna get a, you know if it's going faster.

Scott Stanford:

Sure.

Justin Malcolm:

There's some places where, you know if you don't have this technology, the bus might be going like three miles an hour which is about as fast as you're gonna walk anyway, right? So, and it might be late and you know, obviously you can't afford to be late, depending on the situation. So, if ridership goes up 20%, that has an impact, you know, beyond even the transit system. That's fewer cars in the road. That itself is gonna have knock on safety effects. That's, you know, that's better for the environment. And so I think really, you know, that's evidence both of the driver behavior, but really, you know those kind of the more important after effects, you know, of a system like this.

Scott Stanford:

So basically what you're saying is if somebody's parked in that bus lane, which is usually all the way over to the right side next to the curb.

Justin Malcolm:

Right.

Scott Stanford:

If somebody's parked in the bus lane, it's the bus that will actually take the picture of the license plate.

Justin Malcolm:

That's right. Yeah, so-

Scott Stanford:

Wow.

Justin Malcolm:

So yeah, on the buses that are enabled with this technology, there's a couple of different dedicated cameras that will be on a bus. And like I said it's got an AI computer and connection to the cloud. And so what's happening is as that bus is maneuvering and as it's, you know, completing its route through the city, the camera out the front of the bus is constantly looking at the traffic that's in and around the vehicle, figuring out what the, looking at lane markers, understanding, you know, which, where bus lanes are. It actually uses GPS data to sort of correlate, you know, where it is, whether it's in a bus lane, which lane it's in and whether any of the vehicles around it are in a bus lane. So, you'd be okay. 'cause what happens is if the vehicle which is detected in a bus lane is seen, then the Rekor technology you mentioned earlier actually gets used and actually a second camera takes a high quality picture of that vehicle, the make, model, color and importantly the license plate itself is captured from that vehicle. And that's a surprisingly hard problem. We think license plates are really consistent but it turns out there's a huge amount of variation. And so, the fact that then that a vehicle's been seen in a bus lane and a lot of information about that information about that vehicle's collected, but that doesn't necessarily issue a citation right away. So first of all, a lot of vehicles are allowed to be there. So maybe it's a law enforcement, an emergency vehicle. Diplomatic vehicles maybe you're allowed to be there. But if you're just in the bus lane momentarily, especially if you are still moving, you don't necessarily get a ticket. Maybe people are pulling over to be, you know, because they need to make a, as you said, it's all the way to the right. Maybe they need to make a right turn. That bus, and as you said all the equipment is actually on the bus. That bus has to actually see you in the bus lane. That gets captured and logged, and that information gets communicated up to, you know, to all the robots in the cloud. But you have to be seen again in a bus lane within a certain period of time to get a ticket.

Scott Stanford:

Oh, okay.

Justin Malcolm:

And that could be the bus or a second bus that might come along and actually see you there. And then you say, "Okay, now you've actually satisfied all the criteria to actually issue a ticket." So if you're actually parked your car and leave it there for any period of time, you're likely to get ticketed. But if you just sneak over into the lane, you know, and you're not really blocking or impeding the bus's you know, mission, then you probably won't, right?

Scott Stanford:

Right.

Justin Malcolm:

But, and then, you know, if a ticket is issued, you know, then all the information, you know, it's just a ton of information about exactly where your vehicle was seen, exactly what time it was seen. Again, all this other information about the vehicle itself, high quality photographic information about, you know, about the situation, both from a wide angle context and then very specifically things like license plates. That all gets sent up and an evidence package is created. So, if you do get a ticket, then we probably got you.

Scott Stanford:

Well, listen, I think it's fantastic. The one thing that that strikes me is if they let the mass amount of drivers know that this technology is on the bus it would prevent a lot more people from doing it. You know, rather than, oh, one person gets a ticket, now they know not to do it again. But just to let everybody know that that technology is out there and the awareness I think that would also prevent folks from actually, you know, pulling over to park in those bus lanes. But, just such amazing technology. When you, Justin, when you guys are working with a new fleet, EMS, fire trucks, how do you customize your services to fit unique needs and the operations? Is there, you know, I'm sure there are a lot of meetings that go along with it. You find out what their needs are, and then you guys proceed. But, you know, what's the basic, you know format to how you guys come up with, you know the unique needs and customizing safety protocols for different fleets?

Justin Malcolm:

Sure. So, I don't know if there's a format necessarily, but I think we do, you know, so we are dealing with, you know, specialty fleets. You know, there's complicated problems that we're solving. And so, I think we generally move in with a very consultative approach, right? So, really trying to understand, you know, every customer's different. Fleets operate differently. The context that they operate in is different. So really, you know, what are we trying to accomplish? You know, what are the real drivers of those end goals? You know, what kinds of solutions do we have bring to bear? We've got a lot of different stuff. So, really trying to, you know work with the customer. May have them understand what's possible, 'cause they might not be familiar with that. You know, how what we would offer, you know, maps towards what they're trying to do. But the other thing I, the other factor I think is that, you know, we recognize, you know, often we're putting together more of a roadmap, you know, with the customer rather than, you know, just putting together just something that they're gonna, you know, purchase and install as a one-time event. Because, you know, there's, you know, lots of different factors. There's, you know, budgets, resources available, skills that are available, data. You know, there might be things that, you know, that to really complete the puzzle for them that we might have to augment in our solution, you know, that we needed to put together. That might take, you know, pulling in a kind of data that's not available yet and figuring out how we're gonna get to that. So, you know, often it's, you know, it's really a multi-year program. You know, depending on the size of the fleet, a lot of times you're not rolling something out, you know, everywhere all at once. It might be, if these are large expensive vehicles, maybe it's in new vehicles as part of a new build, like you'd mentioned transit buses earlier. So, it depends, I think on, really on what's, I mean, the key word you said there is just customized, really understanding, you know, what's the right way to do this? What's the right way to, you know, get to the end goal?

Scott Stanford:

Yeah. But listen, before we go, I just want to hit law enforcement agencies for a second. You know, how do they leverage Safe Fleet's next generation in car video systems, of course integrated with Rekor's technology during their patrols? And how's the combination more effective and cost efficient than really the, you know, than the legacy solutions that people used to run?

Justin Malcolm:

Yeah, sure. So again, actually similar set of technologies in a lot of use case, you know, quite different. But, almost all police cruisers these days do have video as well, right? So, it's, you know, along with body-worn cameras, but they have video at least recording the, you know, what's going on out in front of the vehicle. You know, you typically, what's going on in the backseat as well, audio and video. So, in-car video systems that are already capturing you know, everything that's going on with that vehicle while it's on patrol are in almost every vehicle. But, you know, ALPR, license plate recognition systems, historically have not been. So, the legacy systems for LPR are quite expensive. And as a result, they're deployed on a relatively small number of vehicles. And they actually require quite a bit of officer interaction. So, the real change I think we'd like to think of a LPR are everywhere. So you're talking about, again, the Rekor technology, you know, we're looking to put that into every single police cruiser where that's just something that's there by default in the video system that they already have in place. And where that's a just a function of that vehicle while it's on patrol. So, you know, rather than the officer having to very specifically interact with that system and, and try and do a LPR activity itself, just to be have an automatic system where the artificial intelligence is running and being on the lookout all the time, scanning all the license plates that are seen by the cameras as that vehicle goes on patrol, again, capturing make model color the license plate but not necessarily notifying the officer of anything. So that information would be compared in this case with hot lists that are automatically updated by the system which have contained vehicles of interest. You know, so that might be because, you know, be the occupants of those vehicles are, you know, persons of interest to law enforcement.

Scott Stanford:

Right.

Justin Malcolm:

It might be because those are stolen vehicles. It might be because that was a vehicle that was, you know, recently involved in a crime, Or, you know, in the case of the Rekor technology, again, if it's capturing make, model, and color and the license plate, you know, is on the wrong kind of vehicle, says, "Hey, this license plate should be on a white truck, but it's actually on a, you know, a Green Range Rover," then that itself is something, you know, there's an article of interest. And so only in those cases when there's kind of a hit and there could be rules around this depending on the kind of vehicle and the kind of patrol that it's on, but only then does it alert the officer, right? So it's really a passive system until there's something interesting, and then it alerts the officer that there's something there that they need to be aware of so they can take action. You know, one of the most useful places for the officer though, actually is when it's and where we're kind of excited about it is a safety feature is when an officer stops a vehicle, or if an officer sees a vehicle which is stopped in an area maybe at an unusual time, or it's unusual to be there, and they're looking to approach it. You know, they can actually, the system will also notify them, you know, for example, if that's associated with, you know, any of the things we just mentioned. And maybe, that might mean that, you know, it's really not safe or advisable for that officer to approach that vehicle alone. If that was, you know, a vehicle that was used in a bank robbery earlier that day, and there's people in it, and you're alone and it's dark, you know, maybe don't-

Scott Stanford:

You don't want to go near it.

Justin Malcolm:

Maybe just don't roll up and knock on the window, right, because then you're in a movie scene. So, you know, so the really interesting thing there as like I said, is that before these kinds of things would've been in a very small number of vehicles whereas now they can be kind of on every single vehicle on patrol.

Scott Stanford:

It sounds like an absolutely fascinating career to be in, fascinating business. Justin, listen, thank you my friend. Justin Malcolm, director of product management for Safe Fleet. Continued success. And just, you know, as just a layman with all this, you know, not even knowing this is all going on while I'm at home asleep, Justin, you know, just, I'm looking forward, you know, to just seeing what technological advances. I mean, I'm sure there's more to come. I'm looking forward to it in the future. So again, thanks for a couple of minutes, bud. Thank you.

Justin Malcolm:

No, thanks for the conversation. It was a fun chat. Appreciate it.

Scott Stanford:

Thank you, my friend. Have a great day.

Justin Malcolm:

You too. Bye.

Scott Stanford:

Well, unfortunately, that's all the time we have today here on The Infrastructors. Join us next time for our conversation with the Chief Technology Officer at the Central Ohio Transit Authority, Ms. Sophia Mohr.