LPRC - CrimeScience – The Weekly Review – Episode 242 Ft. Robert Brown (Axis Communications)

Episode Date: July 16, 2026

In this episode of the LPRC CrimeScience Podcast, Dr. Cory Lowe speaks with Robert K. Brown of Axis Communications about the evolution of network video technology and its expanding role in retail. They explore how intelligent analytics, artificial intelligence, and integrated surveillance solutions can support loss prevention, strengthen safety, and provide broader operational insights.

Hi everyone and welcome to crime science. In this podcast, we explore the science of crime and the practical application of this science for loss prevention and asset protection practitioners as well as other professionals. Good morning, good afternoon and good evening, everyone. My name is Corey Lowe. I'm the Director of Research here at the Loss Prevention Research Council. And today I'm joined by Robert Brown from Access Communications. Robert, it's very good to have you.

Thank you, Corey. Good to be here. Yes. So you've got a very long background in the industry. You've done a lot. Can you just start off by telling us a little bit about your background, what you've done and how you ended up at access communications? Yeah, I appreciate it. Yeah, it's been a hot minute. I realized when I've been doing this. Started actually in 2002, so I've been doing a bunch of IT consulting and decided to get myself set up at a nice solid, day-to-day job, and that ends up being, fortunately for me, within the IT group at Target Corporation supporting Lost Prevention's first ever deployment of digital video systems. This is in 2002. And it was just luck. I just happened into it, loved the work, was able to see in 2005 the introduction of IP cameras and working through what that meant. Really, and just, on the technology side, building out a lot of the initial video surveillance system that Target used in the early to mid-2000s.

And did that for a lot of years before transitioning over to cybersecurity, spend a little of time there with the Infosec teams. And didn't love it nearly as much as the physical security world. And so I had an opportunity about, actually not about exactly 10 years ago. This is my 10th anniversary month at Axis. We're pretty much for almost my entire time there, I've worked on our solutions management team. So that lets me interact with product development, systems development solutions teams in Sweden, our global headquarters, and the real world. Yes, Corey, honestly, that's one of the things, whether it's consulting or all the stuff we did at Target, I've always felt that technology doesn't exist in a business. vacuum. And if there's a way to do something through process or people, you don't need to

slap technology at it. And that's something they get a chance to do all the time today is actually solve real business problems, real world challenges. It's fun to apply camera technology to it because stuff that I love. And that sort of just sums up everything. We've the last 20-some years. Just trying out new stuff, seeing what works, seeing what doesn't, and learning and growing and just keep keep on keeping on yeah definitely i i something i really like to say is um you know even if there was a silver silver bullet which everyone's looking for you still have to have an operator and you have to know how to use uh use it and all of that so um today we wanted to talk about business intelligence and there's a lot that can be done you know when we talk about cameras if you think

about what happens in retail only a very small portion of it is um you know law of prevention related. So let's just start off by talking about business intelligence with video data. When people talk business intelligence with that kind of data, what are they typically referring to? So it's a really good question. I think, and what I've seen over the years, I feel like this is sort of a catch-off phrase you alluded to earlier. It's, we're using this for non-security purposes. And I've I don't agree with that, but I don't we'll get into more loss prevention type scenarios, but it's basically to me it's translating that raw video, the mountains of data that maybe you don't always use into something that can actually help solve, again, these business problems.

That's actionable. I feel like we've been building those things for a long time now. It's not new. What I think is maybe a new or twist on it and the thought process that I've kind of had for a couple of years at least. I know when I first started, security cameras were really little more than insurance policies for when bad things would happen. And you would archive video for weeks, months, whatever your retention policies were so that you could go easily, go back and find evidence. For me, like the business intelligence is basically everything else other than that. Let's use all of that data, not just for building a case against bad guys, but for helping other parts of our business, which might include loss prevention teams.

and frankly, they often do. Yes. For loss prevention, the criminal intelligence and the loss intelligence is the business intelligence, right? I think you've probably got a thousand different examples of the ways that cameras can be used for loss prevention. What are some of those examples of business intelligence for loss prevention? One of my favorites from my very early days is just an exception-based report is you can tie in your, POS system, you've got a register number, you've got a camera probably above or near that register, and just the simple ability to look through an exception like my cashier has rung his employee

discount for a transaction over $1,000. You know, something that from a data perspective, you maybe have teams running all sorts of interesting exceptions in transactions. you know, and just the simple ability to link that transaction number, it's got a date and timestamp, which the visual clearly has as well, and just be able to play back the video without having to have like two different monitors up and juggle, you know, go back and forth between different systems. To me, that's business intelligence and it helps solve these problems. It's a really simple one.

The more current and relevant, I think everything that you're seeing around ORC and understanding patterns across multiple locations, whether that's LPR, data, facial recognition data, other video sources of data, and other like social media, you know, kind of gathering all these disparate sources of data that aren't just video related to me is another great business intelligence use case that involves video, but doesn't necessarily depend on video. Most definitely. And I think that's the transition that you kind of were just, you kind of just touched on is that it was using like EBR exception-based reporting, right? And that's the anchor, right? So you're starting with the exception and you're looking into it and you're investigating it. Today, there's

all these analytics that kind of turn on its head where you're using the data from the camera as the anchor for a broader investigation, finding some anomalous behavior or finding something that you've seen happen before and you can look into it further. Yeah, exactly. I also like to, just as I've got so many examples of these, but, and I think it'll go into some other, you know, probably more of what we talk about when it's lost prevention. For me, like a time and area or a loiter and a dwell time, like scenario, that's a little bit smarter than it used to be back in the day with just pure pixel changes. But that is such a great solve, especially during COVID. Like when you started doing, you know, a lot more curbside pickup and curbside delivery and just have. having that notification that a car, actually not just a car, but the car specifically with the license plate for this order just showed up. I don't need to wait. That's improving customer service at the same time that that lawyering algorithm or that

timing area says, boy, there's somebody parked in the loading dock that's not a truck after hours. That's all kinds of wrong. So you can kind of apply some of the same analytic or the same, you know, intelligence, but you're solving different business problems. Sometimes it's loss prevention, sometimes it's customer service. Yeah, I love that.

The way that we in loss prevention and crime prevention, we often talk about intelligence-led policing or intelligence-led loss prevention, and it's all about building up the intelligence that you need to do the job. Of course, it gets outside of loss prevention, too, and security. Like I said, pretty... The vast majority of things that happen in retail have nothing to do with crime, loss prevention, or loss at all. So getting outside of that realm, what do you see as some of the great opportunities for business intelligence for the rest of the retail enterprise outside of loss of protection of security?

So quite a few. And I just off the top of my head, I think so much about just counting customers, keeping track of customers as they move through a store. I was at a chance a couple weeks ago to meet with some cities that are doing some things with traffic analysis, intersections and corridors of traffic moving through red lights. And it was fascinated me because the terminology they used was so much like what I'd expect in retail where you've got like arrival on red and they've got metrics where you don't want people waiting at a red light too long because they might try to find side roads or other ways instead

of this main corridor you set up for them and it's like man that's just checkout lanes right there that's how long are people waiting in line are they abandoning that line are they taking too long at the register whether it's self-checkout or otherwise even stuff like this felt like Holy Grail for a while, but I think we're there today where what is my shopping unit? If I'm trying to figure out conversion and four people come into a store and there's only one receipt, that might be okay because that's a family of four, it might not be okay because they're four strangers. But without that knowledge of actually what that shopping unit really is, you're not going to get as accurate of a conversion rate as you maybe could otherwise.

So all that stuff I think is really, really cool and fascinating and such a help for the overall business that is not at all related to loss prevention. Yeah, now this is going to be an unexpected question, but you brought up tracking customers. Yeah. When some people hear that, some people out there in society hear that, they're like, oh, they're like the store is tracking me, tracking who I am and what I'm doing.

Is that what's going on, or would you like to explain a little bit further what we're actually talking about here? That's a great, yeah, so funny, you just don't realize it when you say it out loud. Exactly. The nuance there is, I think it is really valuable to know an individual's journey through a storm. Because we can do heat maps, we can do all sorts of cool things that are all just showing general flow. But there is extra value if it's individual A. But that's all I care about.

Is that in the, I don't care of it. Is Corey or it me? I don't need to know that person. I just need to know that that path isn't necessarily. seven different people that have merged. The data is more precise that way, I guess. So for me, if it's tracking, it's like, I'd love to be able to make sure I know that the same person

passed through all these different cameras. It's not an investigation. In that case, yeah, I do want to know specifically. It's you because I saw you concealing something at the back of the store. I want to follow you personally through the store. But for all these B.I. type applications, these other solutions, it's anonymous. It doesn't matter anything. It is completely irrelevant who you are as a person. Yeah, I run into these questions every single day. It's a lot of fun explaining to people who don't work with these technologies what we're actually talking about.

It's a good discussion, Corey, because it's one of those that I bet if I was maybe on the marketing side of a retailer, I would want to have some demographic data or a better understanding is that a child with that that family that walks in and are they stopping in the kids sections. But that's then become more, what's the process? Do you, if that technology exists, do you want to talk about that as a retailer and say, to the public, this is how we work with this data? And I think it's the responsible thing to do is just that. If you are using this in any way to drive marketing campaigns or to send me specific things because of my previous purchase habits,

I think you've got a responsibility to inform the public that that's what you're doing. And I don't see it happening a whole lot, so I don't think it's something that many people want to touch with a 10-foot pole, to be candid, and just having anonymous data is more than enough for a lot of these folks. Yes, exactly. I think that's a very, I think that's the, that's pretty much the common perspective throughout the retail today. Now, access to cameras, right? You do cameras, but you do a lot of other things as well.

In your mind, what does an ideal retail tech ecosystem look like? So, well, this is another good question. And it's funny, I actually start to think about this as not so much. Well, let me take technology first. And then if I have a second to pivot, I have a second question on that or second answer to that too. And I think it is, for me, just a lot of different cameras that are appropriate for the environment that they're in.

And that's whether you're new at this or you've got people that help you with it. It's simple as, you know, just having the right pixels on target, having the right indoor versus outdoor mix, having pantilt zooms, I think appropriately positioned throughout a store because they never point where they're supposed to when you want to catch something. I like the ability to mix and match some other technologies in there as well, not just the cameras, but like a speaker or a strove or some other ways to kind of alert. And just the fact that all of this lives and plays on the network, I think is important. So, I mean, I think healthy is a good mix of things.

but when I think about what's really like for me, when I like back at my career and what was really valuable to me was honestly as a practitioner, a healthy relationship and partnership between the technology and the business teams and something that whatever technology we were working with, we could figure it out if we were working together. And I, you know, I get, we talk about silos everywhere.

Everybody's got silos. And I don't think you can tear down the science. but you can take field trips occasionally. I made it a point when I was in IT for a number of years. I scheduled a day a week at an empty cubicle where my business partner sat, just so I could kind of be there and listen and hang out. There are silos within the IT department. So we would make a point of bringing in people from like the networking teams and the cyber teams

and the server, workstation owner teams. So it's funny because that healthy ecosystem to me is more about the relationships because that then allows you to solve out-of-the-box problems because you guys are aware of whatever else is working on. Yeah, I think that's one of the lessons that King Rogers, I'm pretty sure you're very familiar with him, really taught me was the importance of every time you have that opportunity to engage and interact with a business partner,

whether that's someone within your department or between other departments, that's a key to innovation. You know, it's one thing to be able to make a business case and formally say, here's what the projected ROI is and all of those things. It's another thing entirely to get that buy-in from the very, very beginning so that when you pitch that business case, everyone's already, you know, thinking the same, aware of what you're trying to do and where you're trying to go. and also see that as a common vision. Yeah. Yeah, so I'm a quick anecdote, just as an example that came up as I was thinking about this.

Because I took the time to kind of spend time with business partners, this is way back in the day, it's going to date me even worse than early IP video. It was like when Black Friday used to be a crush of people. And there were legitimate safety concerns. And when we were kind of rolling out this IP video, we knew that a lot of people knew how to get to and look at video. And we were worried about the loss prevention team's ability to actually leverage that video effectively when we had potential looky-lose on the network trying to find how busy stores were.

And so we just talked about that. And I had worried that my business partners were going to try to find some, weird technology solved for a thing that felt pretty unsolvable until we realized that it wasn't video that people wanted. They just needed like a snapshot. They needed some way to take a look at what things looked like. So we together built a supportable and supported application around Black Friday that allowed folks in the field to get a glimpse of what the store looked like without trucking that heavy bandwidth across the WAN and killing the network for, what's probably most important on Black Friday is those transactions making sure people are buying stuff

not looking at video of the store down the street they want to see if it's doing okay. Yeah. What? Retail is about selling stuff? Yeah. Yeah. And so again, it's a long answer, I think, on that healthy retail technology ecosystem. But for me, it goes to, it's the relationships, it's the partnerships. And whatever else you do, technology-wise, you can figure it out if you've got those things in place.

Everything beyond technology makes the ecosystem. The sensors and all the tech is important, but people in process is critical. I was talking to a colleague the other day, and I was expressing my feelings about artificial intelligence sessions at conferences. It always feels like a lot of those sessions are all buzzwords, and just a lot of, a lot of buzzwords more than anything that's actually useful. So one of the things I always like to do is try to kill as many buzzwords as I can. I really want to go deep with these things.

And one of the things I think you and I both are interested in are open architectures. It's a buzzword that you hear out there. Can you explain to the listeners what an open architecture is? You know, what does it mean in practice and what does it not mean so that we're When they hear that word, they can identify whether someone actually knows what they're talking about or not. All right. So dad jokes aside, I was thinking about like it's just a roof without a roof. It's an open architecture.

So to me really what it actually means in practice, I guess, is I just like to think of all these different tools and technologies as we're all living in the sandbox. Right. We have everybody's got kind of their own definition of what that sandbox looks like or that playground looks like. And whether that's, again, a little bit in the weeds, like it's all a Linux shop or it's all a window shop or it's all web-based or it's all on-prem. Whatever your definition of a sandbox is is sort of besides the point, the open architecture is if you've chosen that, you should be able to work with other tools and other technology in that place that you've defined. That's the openness of it. I shouldn't have to change how I play in my playground just because I want to use your technology.

And to me, that open architecture, in many ways, just whatever parts and pieces you want to as a retailer want to work with, you have the freedom to do that. You choose them. And they'll all still work well. And they'll also maybe work well together, but not necessarily always. Yeah. as opposed to a closed architecture. So being open and the ability to integrate things with as little friction as possible and closed as not being able to integrate technologies together and work with them.

It's an important concept to me because from an intelligence-led loss prevention perspective, we talk about building intelligence as in you have to collect all your data points and you have to bring those together. And as you bring more data points together, you build the context around what's really happening. Until you get to the point where you have intelligence, something is actually actionable because you know the full context of what's going on in a place at a given time.

That requires that you bring in those data points from different sensors. And if you can't connect them, that can create challenges. So that brings us into the costs of closed systems. So when we think about open versus closed systems, what are the biggest risks and limitations of closed systems? What does that mean for the retail enterprise and what they are or are not able to do? Yeah, so it's funny. I feel like this term may have fallen out, but I go back to when I, you know, there's one throat to choke. There's benefit and value people like this closed system is like, look, I don't care.

I got one guy I can go to when there's ever a problem. Where you have the risk and constraints, I can think of three right away that are really impactful. And one is probably a whole separate podcast on lifecycle management and how difficult it is to plan and budget and figure out if I've got devices, we'll say cameras, it's my closed ecosystem in a ceiling that need to be replaced. like how do I time that and plan all that? That becomes a lot more challenging in the closed system because your choices are limited.

You are not typically going to be able to upgrade or lifecycle beyond what your clothes manufacturer provides for you. And in fact, you may actually need the lifecycle before you are ready to lifecycle, depending on different things within that manufacturer. So that inability to kind of really, really own the life cycling is part. of it. The second part to me is that it means you can't really innovate at your own pace. It's a lot harder to dabble and experiment and try new things and to say we're fine with 95% of our camera infrastructure, but we feel like it 5% and some key markets where we're seeing some exceptional shortage. We want to try something different that does not adhere to this

closed ecosystem. I want to go to a trade show and go to the startups, right, and work with those folks. that you can't pull that into that closed system, you're going to be less likely to try those innovations. And I guess then my third is similar to that. Like you might want to build stuff in-house. You might have really smart, really creative coders. You might have people that are, you want to have this proprietary solution that you're building is going to be a competitive advantage.

Close systems don't even allow you the option to do that, to build your own systems or solutions or reports, whatever you have to save the day. So I guess all of it kind of when I say that loud, Corey, it boils down to like there's no freedom. You don't have the ability often in a very close system to be able to do the things you want to do. You've got to go with the recipe book or the tools that I provided to you by that, the closed system. Yeah. And I thought it was fascinating that you actually started off with one of the benefits of the closed system, which is, you know, the ease of managing it in some way.

right, having that one-throat trope and being able to do that. You know, being strategic as being able to look at all the pros and cons of the different types of systems and make a informed decision that is best for your business. Are there any other, you know, benefits of a closed system? Benefits, the one's for sure. And I think on the life cycle, part of this might be, again, a little bit old school. when we first started at this, we were buying NDRs at Target. That was the solution.

And the VMS ran on the same boxes as the CAT-6, sorry, my co-act cable would run in to the back to convert that analog signal to digital, which was the same box that ran the VMS. So at a very simple thing, we were looking for cheaper storage. And we couldn't spend money on cheaper storage because that was bundled with that box. You know, so to me, that's when you say life cycle management. There may be other pieces to it, other factors besides just the devices that you're buying or the software that you're running that you maybe don't have a choice to life cycle if you've got a closed system. I love that example, by the way, because that really, it really constrains what you can do with those budget dollars, right?

So if you are stuck with a specific component because of a partner, that you have less freedom like you said earlier. Yeah. What else are you about the same? Sorry. I would just say I think the bigger the enterprise, the bigger you are, you do have purchasing power that maybe some other companies don't. So that's one of the things that you get a chance to see more of

when you are looking at pricing out things like, of servers or switches or storage arrays. To me, the bigger you are as a retailer, the better an open system is. The smaller you are, frankly, a closed system can be really appealing and really effective and honestly not have any of the constraints or limitations that I bring up. You're not going to build stuff in-house because you don't have a development team in-house to do it. You're not worried about lifecycle management because it is easier if somebody else

thinks about that, especially if it's like a cloud solution and there isn't really a need for non-prem storage or whatnot. So to me, that's sort of where I see the sort of dial of the bigger you are, the more important it is to be more open. Yeah. A second buzzword that I love to hear is edge computing. And you hear people talk about near edge, far edge, all of those things. Very edgy.

Yeah, very, yeah. But I'm bump. No. But edge computing makes a lot of things possible, right? It really changes how you actually set up entire networks of cameras and entire, you know, networks in general, not just what's coming from the cameras. But, and that's because of the data and how it's transmitted. So can you tell us a little bit about edge computing what that is and what that, what it can do for real-time intelligence,

making decisions in the moment. Yeah, it's great. I'm glad you used to turn far-edge because that's kind of where we see ourselves at access. It is, because the edge used to be the cameras. It was just stuff in the ceilings. I think a lot more commonly now, you'll talk about that, like as an appliance or a bridge

or something that may be on-prem at a store, but that allows for communication with a cloud solution or whatnot. So that, anything sort of, I think out of the store, out in the business areas is going to be your edge. On the far edge where the cameras are, the biggest benefits of any kind of compute out there is that there's no latency. There's literally nothing to separate the video stream from, let's say, in analytics that's running or image processing or whatnot.

So there's a lot of power and a lot of benefit into doing as much as you possibly can on the edge, because even though bandwidth is a lot cheaper than when I started in this business, it's still expensive. And when you can do really fun and creative things like let the camera do a lot of heavy lifting and say, here's an image of a scene, and all I want to do is ping a large language model,

a Gemini or a chat GPT or whatever, and say, give me more details about that, which is really, really cool and really powerful, those transactions cost. and the bandwidth to send video costs. And there's still expense there. So for me, what I think is really good is have the best of both worlds. Let the camera do as much as it can.

Let that far-edge computing take care of a lot of things for you in real time, right? And for a lot of the business, a lot of these BI applications we're talking about, the zero latency, the lack of, you know, compute, all that stuff is less relevant because you're looking for this information after the fact. So I think the more real time it is, the more bandwidth-intensive intelligence is take advantage of the camera's ability to do some of that work without requiring on cloud or data center or other compute. Yeah. And those two, two concepts between open architectures and edge computing.

I think they have a lot of implications for what is possible, not only in retail, but many, many different industries. What do you see the future looking like with these computing at the edge and then having open systems where you can bring in a lot of data points and add this context to retail and beyond? Yeah, it's so good. There's so much about within loss prevention is you're looking for patterns. You're looking to shift through mountains of data that maybe doesn't seem like it's connected to help solve a case or to help get somebody, especially these really, you know, some of these bigger, more complex things.

Obviously, if you've got a guy's face and he's, you know, taking stuff. But I think that between generative AI in particular, boy, I see just huge wins and like automating. like case management and case building and some way to take out some of the time and effort it requires to put together. Like I've got video from 20 different cameras across three different stores across, you know, two different states that I need to go out and take the time to find. And especially, boy, if I'm an investigator and I'm not this familiar with a store layout, so I don't know how easy, how to easily track that person through a store based on their different

cameras. Like, I think there's just so much great possibility there for the, for AI to potentially be able to, um, gather that video evidence, stitch it together if, if necessary to tell the story to, to transcribe, maybe just describe what's happening, you know, in that scene. If there's a body worn camera to have, uh, natural language, transcript generated, like, all of this stuff together. Um, and it's funny, even as I'm talking about this, Corr, and I, I know you've been doing this a little while too. I go back to fundamental challenges that have been facing retailers for at least 20 years. And I talk about this a lot when I present on this stuff is we don't have enough time, enough budget or enough resources to do the stuff we need to do. And I think AI has the

potential to help with all those things. It's just all of it. I think it's going to be really, really interesting to see where it goes, you know, while maintaining and balancing privacy concerns and legal considerations. You know, I think we talked like even just LPR, again, a whole topic for a whole other day. But there's a lot of value in the data there as long as it's done in a way that benefits, I think the last prevention teams without necessarily all these other privacy concerns that can stem from improper use and sharing of that type of data. Yeah.

And I really love spending time with you because you were just so aware of all the practical challenges of making some of this work. I love the discussions we've had because most of the times I don't know some of those practical constraints. I know a lot of the possibilities, but then you're always one to say, oh, well, you think about this too. How do you do this at scale, for example? But you've been around for a very long time. And one of my favorite questions to ask on this podcast is kind of the timeless strategic insights that people have to offer. So I just want to wrap up the conversation today and ask you, what leadership principle or value in business have you carried with you throughout your career that you believe has made all the difference, has allowed you to build partnerships? and really be successful in working with others.

That's a great way to close here, Corey. I appreciate talking with you, too. This has always been fun throughout our interactions over the years. For me personally, I think the first one that comes in mind is curiosity. I've always been curious. I think that there's always opportunity when looking at anything to look at it from a couple of different angles, to give a couple of different perspectives, to be open, I guess,

to the possibility that my idea isn't the best idea or the right idea. It maybe is close to it. And so it's curiosity. It's honesty with yourself and with what you know and what you don't know. And then I throw trust in there as well because trust, I think, applies to a lot of different things. But trust that you can, you know, through those other two pieces, kind of get to the answers.

that make the most sense and learn from mistakes if and when you make them. And it's so fun. It just feels like it's just a silly thing. Talking about kids earlier, and I think that I remember with my children is really working on this idea that there's not even a mistake, it's just opportunities. And the way you learn is just by trying stuff and figuring out what works and what

doesn't work. And one thing I'd hope to impress on my kids, too, when they were growing up, was just that curiosity and willingness to just try something, even if it doesn't work, because that's how you learn for the next thing you try and the next thing you try. So that's why curiosity for me comes down to it all with everything we're doing and just a willingness to roll up your sleeves and be humble and work it out. Thank you very much, Robert.

Curiosity, building trust, and seeing mistakes. as opportunities. I love that. And I thank you very much for spend some time with us today on the LPRC Crime Science Podcast. I always learned something when I spend time with you, and today was no exception. So thank you very much. Thank you, Corey. Thanks for listening to the Crime Science Podcast, presented by the Loss Prevention Research Council. If you enjoyed today's episode, you can find more crime science episodes and valuable information at LPRsearch.org. The content provided in the crime science podcast

is for informational purposes only and is not a substitute for legal, financial, or other advice. Views expressed by guests of the crime science podcast are those of the authors and do not reflect the opinions or positions of the Loss Prevention Research Council.