DigitalThreshold_LIVE_Transparent-1

Digital Threshold Live is a series of webcasts featuring new guest speakers in each episode from a variety of industries moderated by Evolv Technology's Co-founder & Head of Corporate Development, Anil Chitkara. Learn More.

 
 

Episode 4:

Why Technology Convergence in the Digital Threshold Matters

Took place on December 3, 2020 at 1pm ET

Watch the OnDemand Version below.

 
 

Can the digital transformation of physical security deliver an intelligent security platform - something that’s more than the technology itself, something that has practical application?

Join us as we talk with Mahesh Saptharishi, CTO at Motorola Solutions, about the technological possibilities at the intersection of sensors and AI, exploring the business drivers, the technology and ultimately the effect on humans. Mahesh leads innovation across Motorola’s platforms in mission-critical communications, video and command center software.

Our conversation will explore his thinking about next-generation solutions that make communities safer and help businesses stay productive and secure. From advances in AI and data-driven technology, trends are emerging. The nature of access control is no longer just a function of identity, it now includes contextual elements. And, security and safety can be considered distributed; rather than singularly focused in a fixed position, multiple roles engage, collaborate and respond. Dynamic situational awareness requires a change to contextually based access.

Don’t miss the opportunity to hear this thought leader’s perspective on trends in this rapidly transforming space.

Motorola Solutions is a global leader in mission-critical communications and analytics. They make communities safer and help businesses stay productive and secure. Motorola Solutions is ushering in a new era in public safety and security. 

 

Dr. Mahesh Saptharishi

Chief Technology Officer

Motorola Solutions

Episode 4 Transcript

Anil Chitkara:

Welcome to the Digital Threshold Live, here we bring you professionals and practitioners at the intersection of venues and technologies. They'll be sharing ideas and inspire you to how to make your venues safer and more enjoyable for your visitors, your guests and your staff. Today, I'm very pleased to welcome Mahesh Saptharishi, CTO of Motorola Solutions. I've had a chance to talk to Mahesh few times, and he is going to be very enlightening for all of us today, we've got a lot of really good topics to cover. Where I want to start though, is Mahesh, we could equally be talking to you about the COVID vaccine today as we are about AI. So I'm very interested if you could share a little bit of your background and how you steered away from something that might have had you talking about COVID and more talking about AI and analytics.

Mahesh Saptharishi:

Sure. During my early days of my career, I was delusional and I thought that I really loved biology. And so I actually spent some time at the National Institutes of Health, working on immunology, and got into the mathematical modeling for immune systems and then realized that I was really good at killing cells and wet science was just not for me and eventually decided the mathematics was actually kind of more interesting. And so I went more into computer science and mathematics side of the world as opposed to the biology side of the world. So yes, I did spend a fair amount of time in immunology, I know enough to be dangerous, but not enough to be competent.

Anil Chitkara:

Great. And so are you generally more hopeful now than you were a few weeks ago with all these recent announcements about the vaccines and the performance in the trials?

Mahesh Saptharishi:

Absolutely. I think it's incredible how these multiple entities, especially in collaboration with the National Institute of Infectious Diseases, has done such an amazing job in understanding the nature of COVID, the messenger RNA based approaches at Moderna and some others have pursued. Super promising, and I think I'm certainly very hopeful that this is a good outcome for us. Unfortunately, by the time everything rolls out, it'll be close to the middle of next year, I think. But that said, this is the underlying technology here, I'm speaking to my bias of technology here, but the notion of mRNA, messenger RNA based vaccines being effectively software that can reconfigure our immune system to respond to diseases like COVID. And this being a hopefully an approach we can replicate as future pandemics or future potential pandemics come to the forefront, I think to me, that's very comforting and I think we all should be comforted that hopefully we've learned lots of good lessons from this really hard time.

Anil Chitkara:

Right. Yeah. And it's been interesting, as some people have said this seems to have happened so quickly. But I think a little bit to your point is, there's been a lot of work on the platform, essentially the technology platform and this vaccine sort of sits on that platform.

Mahesh Saptharishi:

That's absolutely true. That's absolutely true. This is... Things that seem like they take a minute to produce usually have years of preparation that go into it in advance. So I think that the fact that the machinery that produced this can be reproduced for other things as well, I think that is the compelling thing here. And also, obviously very promising for the vaccine like for COVID.

Anil Chitkara:

So I'd like to stick on the technology area, maybe transfer out of the wet lab into the zeros and ones a little bit, which will make both of us more comfortable I'm sure in that domain. There're so many different terms out there, there're so many different technologies. There's AI, there's machine learning, there's computer vision, there's big data, there's analytics. My head spins frankly, I know a lot of our head spin. Can you just untangle it a little bit, what are some of these things? And what is the difference among some of these things? And how should we think about them at least as areas of technology?

Mahesh Saptharishi:

Sure. And I think if I were to propose some sort of formal definition it'd probably be quite contentious, at least in the academic and perhaps even the industrial world of things. But roughly where there's broad agreement is that machine learning are the core algorithmic capabilities that power AI. And artificial intelligence, intelligence is really what happens during the execution of a system. So effectively, when cameras or when systems see things, detect objects, respond to what the objects are doing in the scene, that is artificial intelligence, but that ability to detect and the ability for that system to adapt to the environment is powered by machine learning algorithms. So a way to think about it is artificial Intelligence is potentially what we do, what we act, machine learning are all the chemicals in our brain that are making it possible, the neurons firing the interconnectivity between elements in our brain. The formation of that interconnectivity is effectively powered by machine learning. And so that's... At that level, that's how I separate those two things. Big data is more, I would say as the name suggests, it is the thing that AI applies to.

Mahesh Saptharishi:

So this thing, the bigness of this data is really characterized by the volume, the velocity and the variety of data that hits any artificial intelligence system. So it's more of a characteristic of what AI can be applied to, versus AI itself. And so I think that's how I would characterize big data. And lastly, you asked about analytics, I think analytics is really the outcome. So somebody who's interested in business intelligence, who wants to understand a customer's journey through a facility, asking that question and the analysis that goes behind it to be able to answer that question through visuals, through dashboards, through any sort of set of metrics, the derivation of that, that's really analytics. That could leverage AI, but is the outcome of that process.

Anil Chitkara:

That's very helpful. That's probably one of the most compelling and logical for me at least, explanations we've seen. So thank you very much for that. So you spent a chunk of time at Carnegie Mellon, academically exploring these areas. And since then, you have spent time professionally, essentially productizing the technology in the areas. Can you describe how these technologies are manifesting themselves in the real world with real customers? There's so many different academic aspects and deep technical aspects of them. But what's happening with them? How are they manifesting themselves in the real world environment?

Mahesh Saptharishi:

Sure. So, I think AI, especially when it comes to understanding audio in terms of speech, understanding video of all kinds, has been evolving now over many, many decades at different levels of performance and capability. Probably in the past 10 years, one of the key enablers has been the development of a new, perhaps not to call it a new type of processor is not the right word. It is really an application of something that existed before GPUs, graphical processing units. But their capacity to do large vector operations very quickly, transitioning that from the world of graphics, from the world of gaming, from the world of having a very powerful display and a user interface, that moving into the world of AI where many similar computations exist, has taken something where typically these algorithms that were quite powerful, but were not real time in nature or not even close to being real time, suddenly becoming something that could actually operate in real time, or those that were very computationally intensive, even if not real time, because of the intensiveness of what was required there, them being not practical for many applications.

Mahesh Saptharishi:

GPU is now enabling that more widely, I think, has made this quite powerful as well. And so a lot of the algorithmic developments through the 90s and through the early part of 2000s, were at one point, neural networks were considered to be yesterday's technology, suddenly saw a huge resurgence because of GPUs. That combined with the availability of data. So machine learning, the performance of machine learning is directly proportional to the data that you have access to. And I think storage becoming cheaper, network bandwidth becoming cheaper, the ability to collect data becoming more practical, that acted as a fuel that powered all these algorithms to develop further and actually reach their performance potential and becoming practical through the processor technology that has come out.

Mahesh Saptharishi:

So now, what used to be a speech recognition solution that used to require multiple Intel Xeon processors to run, you can whip out your iPhone or you can rip out your Android device and speak to it with dictation, and a lot of that computation can actually run right on that edge device and be real time. You can leverage AR as an example through the camera on your phone and that can work in real time. Those same processors now run on cameras on IoT devices of various sorts. So increasingly, what we're seeing is that there's this ambient intelligence that is now being powered with AI largely because of this new processor paradigm that is leveraging an evolving set of AI tools that have come up in the background.

Mahesh Saptharishi:

So all the things that we're doing on our phones, all the things that we're doing when we log into our bank, when we use our credit card for fraud detection, there are so many elements of our everyday lives where that data are being analyzed on a regular basis, anomalies are being determined, things that could affect personalization to give us a better sense of service, all of those factors are things that are around today, have evolved over the past many days and it's ambient, we don't particularly pay a lot of attention to it, but it's all AI, it's all machine learning. And that's the substrate upon which we're building a lot of other things, including camera technology and such.

Anil Chitkara:

Right, right. Interesting. And the question I get a lot is, when we talk about our products and having AI in them, is it replacing people or is it helping people? So as you think about the use of AI, certainly, software and other technologies have been to automate certain things, some make processes better or faster. But when you think about AI and that set of technologies, how do you think about it relative to its impact on people?

Mahesh Saptharishi:

So I think what AI today is very good at is taking mechanical tasks that are perhaps complex in and of themselves, but really mechanical in nature and being able to do it in a way that is more efficient than then us humans would be able to do. And so think of it as counting the number of people who enter a building. It used to be that someone at the door stood with a clicker and kept clicking, counting the number of people who came in. That is something that's perhaps done more effectively by a sensor plus an AI solution of some sort. There are diagnoses, for example diagnosis of X-rays, radiological applications, which are more complex than counting. But it is actually mechanical, where you're checking for certain sets of things and provably today, AI based systems can actually reach a higher level of performance and many radiologists get along on multiple different fronts. And that gives you two I would say, very different categories of people with different skill sets, different levels of skill sets, both of whom jobs are either augmented or in some ways actually replaced by what humans do.

Mahesh Saptharishi:

But that said, humans are not static entities in terms of how we apply our intelligence. And artificial intelligence being able to augment a lot of the capabilities that humans have, also, I think, is allowing us to focus on different sorts of problems, problems that require keeping capabilities that today AI does not have and as AI perhaps encroaches into those capabilities as well, I consider human evolution to be something that also progresses along the same way. So at the end of the day, AI with proper design, I would say gives us a perspective into the world, augments our own capabilities, where ultimately, I think of AI as augmenting human capabilities and opening up new things and roles in the job market that do not exist today, while certainly certain jobs will probably go away, certain jobs will change in some manner. I think augmented intelligence with AI is going to be something that increases the number of things or changes the number of things that we humans will probably pay attention to. And all hopefully for the better.

Anil Chitkara:

Yeah, I think that's interesting. And if I just take it down to some practical, direct conversations I've had with customers about their security organizations and what are they doing with them, there's definitely a sixth sense that security professionals have, whether it's former law enforcement or whatnot, when something just doesn't feel right. That seems like something that's hard to automate, but if you take the other side of it, which is you're looking for a child that's lost with a red jacket and a black backpack, if you can, instead of having a team of people slowly looking through video, have that automated, you can use AI and technology for that and redeploy on those higher value areas. Is that something that you're seeing, or you're thinking about?

Mahesh Saptharishi:

Oh, absolutely. I think... Motorola has this long legacy of mission critical communications. We have law enforcement, we have first responders of all kinds, people who work in enterprises, entertainment venues, stadiums, et cetera, who are all in the field, who are doing their jobs while they are actively doing many other things. And we see the connection between things like video and audio in a way where the capacity of AI to do the looking for you, focus human attention to say, "Okay, whatever is happening here is anomalous potentially, there's an individual who's in a location that they're not supposed to be, they're carrying something, perhaps that they're not supposed to be." And that acts as a tap on the shoulder of that person in the field to say, "Hey, here's an early warning for you, perhaps there's something that you're not... It's not in your field of view right now, but warrants your attention."

Mahesh Saptharishi:

That allows that person who's doing many things at the same time to do everything as well as they possibly can, by taking that very monotonous task of just watching, observing and then calling out just the things that require attention, taking that and automating that piece. But once that attention is focused, it is up to human intelligence that now determines whether that requires action or not. And we're doing that across the spectrum. We have a solution called radio alert, that effectively connects what humans are seeing and taps the shoulder of the person who's carrying that radio and says, "Hey, mister security person, please look here." Because this requires your attention, maybe it's something very benign, but maybe it's something that requires some action. But that is a decision that that human needs to make. And that's a decision that that human can make without actually having to stare at a bank of video monitors or whatever else.

Anil Chitkara:

So very much as a consumer, I use my phone to see my calendar, to make a phone call, to enable the things I'm trying to do. You're talking about weaving technology into people's processes and augment as you said earlier, what they're trying to do.

Mahesh Saptharishi:

Yes, I think we're entering an event driven world. So it's not the age of watching TV and Stumbling on something that is interesting for you to watch, it is the age of searching for what you want and watching that very precisely. Or getting an alert on your phone saying, "Ah, something is now up that may interest you go watch this." We live in that world today. It's an event driven world today. And I think security needs to evolve to that event driven world as well.

Anil Chitkara:

Right. So there's a lot of different technology we've talked about, we've talked about some use cases or applications, particularly in the security area and a little more broadly. There's so many different places you could apply the technology, there's so many different ways to productize it for people. How does your organization figure out where best to apply this technology and what it specifically needs to do?

Mahesh Saptharishi:

Absolutely. I think the so often, especially in the AI world and in the IoT world, we get trapped by this notion that, "Hey, we have this really cool tool, let's go find a problem to solve with this tool." And that often leads to very suboptimal solutions. The way we approach the problem is very much a design lead approach. And what I mean by design lead approach is we actually have a human factors team that job shadows our customers and really watches what they do, interacts with them without necessarily affecting the way they're going about their day to day job. But really constructing a customer journey. And across that customer journey, we map out what the jobs to be done are. And for each of those jobs to be done, we ask the question, what is it today that that person may just be doing very mechanically or perhaps is doing as part of something that they've done all the time and hasn't really thought of it as a problem.

Mahesh Saptharishi:

But if you now make it better, you simplify it, you automate it, you change some of the things that make the job hard, how can we make that significantly more effective, increase the performance of that individual in that job to be done more effectively? That is the design aspect that we extract from that human factors problem. So identify the bottlenecks, identify the inefficiencies in what is done today, take that customer journey and optimize the job to be done through the course of design. And through the course of design, ask a set of what if problems? What if we could apply an AI algorithm to do this thing that requires some amount of sensory intelligence, perhaps or analytical intelligence, but largely is something that can be learned with a corpus of data that we can collect or already have access to?

Mahesh Saptharishi:

And now, by virtue of us automating that sensory action, or that analytical action, can short circuit this process, this workflow that this individual's engaged in and give them that time necessary to either do things faster, or do things earlier, and if it's a risk mitigation, so you should be giving them a greater runway to respond appropriately to that risk.

Anil Chitkara:

Right. I find early in our days, when we were designing our products, I did a very similar thing. I went out, I sat side by side with customers, and I think about it as customer empathy. And there's things that happen on the ground, at the location, whether it's a security professional or guest services, or whomever it is, is interacting with their visitor. And you can't just take a picture of it, you can't just write it down, there's an emotional thing. There's something more there, and you have to be at the site and feel it. You need to be talking to the security professional 45 minutes after they've been having screening people coming in and they're tired and their fingers tired of holding the button down on the hand because they've been doing it for 45 minutes.

Anil Chitkara:

And those are the things that in a traditional survey, or question and answer, you're not going to get insight into that, but when you put yourself there in their shoes and have that level of customer empathy, you really understand some of the more visceral areas to create value. And I think that's where I see a lot of adoption at a human level taking place.

Mahesh Saptharishi:

Absolutely. And we've talked to lots of physical security customers where in some ways they describe part of their job as long periods of boredom with moments of terror. So it's like it is one of those things where, when they do have to respond and react in some way, it is a stressful situation. The time window required to react ends up being very short, there's a decision that needs to be taken within that time window and the cognitive aperture tends to reduce very rapidly in those times of stress. And so when you have a tool that you're offering them and you want to figure out what that tool is to simplify their life, it is that making sure that the data that they're able to decide, used to make a decision is really and truly something that they can consume with that reduced cognitive aperture without distraction. And it's something that actually leads to good decisions ultimately for them.

Anil Chitkara:

Right. You were talking about the building technology to address some of the work that's being done by these people. One thing that I find interesting is, it's fairly straightforward to automate something, but to go a step or two further into something that may not be articulated by a customer's need, or maybe so visible, I think is courageous and necessary. And certainly in the productization that you've led over the years, you've done that. There's a lot of products that were first up. So how do you figure out how much further to push it and that that does align with where the customer is going to need to be or would want to be?

Mahesh Saptharishi:

Yeah. I think a lot of it is figuring out the changing landscape of the job that that customer is engaged in. The job that that customer is engaged in today is going to change one way or the other over the coming years. And I think what we want to optimize for is where that job is going. But really align that with what is the changing nature of tools available to help that person do their job? And oftentimes, those two are thought of separately and they really aren't that separate. And I think you need an understanding of the customer, understanding of the job that they're doing, and you also need to understand where technology is going. I think 10 years ago, if you had talked about a lot of the AI algorithms that are available today to automate a lot of decision tasks, a lot of the analytics that are available today, the performance probably would have been one where people said, "Ah, not necessarily, not really useful to do." Perhaps the cost is not at the right place, et cetera.

Mahesh Saptharishi:

But I think by tracking the future of where technology is going, we were able to start thinking about, okay, there is going to come a time where if we meet these types of performance criteria, and this type of cost criteria, then applying that tool to simplify the user problem, or even change the nature of what that person does, previously, they might have been spending a lot of time doing something, but that something doesn't really deserve their intelligence, it doesn't really deserve their attention, they could be doing something else. The changing nature of their jobs is actually enabled with technology overall to make the things that they need to do, the overall objective of what they need to achieve, in many cases security, it's not really effective response that's the core objective, it is prevention. You want to prevent the bad outcome in the first place.

Mahesh Saptharishi:

And ultimately, the customer role change, I think, going into the future with physical security, is less about not how you investigate, how do you effectively investigate after the event has happened? It is about how do you respond immediately to mitigate the risk effectively, but now increasingly, how do you take steps so that you can prevent that bad outcome in the first place? And the job role is changing and we're effectively leveraging AI, leveraging tools, leveraging products, sensing modalities, new sensing modalities, to really move that job from investigation to prevention across that continuum.

Anil Chitkara:

Right. So we can flip the balance from response and react to more prevention, right? And that's just a little bit of using technology to do that. So certainly in the last eight or 10 months with COVID and all the impact it's had on people in the economy and businesses and locations and schools, there's been certain trends that I'm sure have come out of that. As you've thought about where to go, as you're saying with new products, what are some of the trends you've seen that have come out of the current environment that make you think about security and technology and products in the future a little bit differently?

Mahesh Saptharishi:

I'd say the first couple of trends I think, would not be news to the majority of the audience. The first couple of trends are, there's more of the watching and the observation that is now going into the realm of AI doing that job versus humans doing that job. The second is I think, the nature of the sensor modalities that are available today. I think when we started our conversation here and we were talking about big data in particular, its volume, velocity, variety. And that last part variety, when it comes to physical security is really starting to increase in the sense that it's not just cameras, it's not just motion detectors, it's not just a metal detector, it is more than that. Sensing modalities that give you richer information, but the variety of sensing modalities, when you combine some of these technologies together ends up becoming a more powerful solution. And I think that is a key trend on the sensing side that's progressing.

Mahesh Saptharishi:

The third trend is really all of this really cloud connected in many ways, and it's not the cloud that's the buzzword in that scenario, it is really giving customers without giving them the burden of having hosting infrastructure locally, giving them the benefit of all the processing that they would need to run capabilities to get insights, to get the performance out of the system by combining this variety of data sources to solve whatever problem that they're trying to solve. So if it's prevention for physical security, it is being able to integrate all this information, but endogenously within the facility and exogenously from outside the facility, all in service of making sure that their risk is reduced. That connectivity, the accumulation, aggregation of information, the processing of it, to make sure the security lens of that particular customer is now put in place and the right people are alerted or notified abroad into the loop to become part of that risk mitigation process. That cloud connectivity is I'd say the third really significant trend that's happening here.

Mahesh Saptharishi:

And the fourth trend that I think becomes quite interesting here as well, sort of dovetails on the last one and benefits from the first two, is that historically when you start a physical security, whether it's CCTV, whether it's access control, CCTV the very acronym is closed circuit television. It is really islands of information, islands of sensing that operated independently from each other. Those islands are disappearing, and along with those islands, the interconnected web of sensing in service to whatever job that our customers are trying to do, that web is now bridging between the private and the public side of the world. So public safety is becoming more tightly integrated with private security, enterprise security. And when you bring those two things together, your entire response workflow and your prevention workflow, and even your evidentiary workflow after the fact, all become highly optimized with people, human intelligence applied more appropriately on things that really deserve human intelligence and all with the hope of now creating a better situation, a better risk outcome for our customers.

Anil Chitkara:

Interesting. And so we have these four trends and areas of technology development, as we think about a venue and sort of bring it down to a threshold or an area, what's happening there? So we've had access control for a while, we've had identity solutions for a while, we've had certainly metal detectors for a while, but what's happening at that level, what are the changes you're seeing at the customer at the threshold?

Mahesh Saptharishi:

Yeah. So I think even probably the thing that's top of mind to everybody is COVID these days, right? So it's like pre COVID and post COVID is probably the two eras or epochs that we can think about. Pre COVID, it was all about getting customers from one side of the venue to the other side of the venue, where they could enjoy the venue, improve customer experience, improve user experience within the venue most effectively, while making sure that they're safe, while making sure that they're secure. And so it was really making sure that you're not creating bottlenecks, you're not creating situations where people are frustrated waiting in long lines. Post COVID, it's all about in addition to that, in addition to all the security and the safety concerns, it's also about making sure that you're creating frictionless, touchless experiences. It is about making sure that you're not creating a circumstance where people are congregating where it is unsafe to congregate, where you're solving one security problem, but at the same time creating another one as a consequence.

Mahesh Saptharishi:

So I'd say that's the first bit. It is this notion that I think, along with this notion that you need a high throughput solution, the threshold cannot become a bottleneck. Increasingly, it is also about making sure that there is the safety element of it, where you're not sacrificing in venue security for out of venue risk. So I would say that's the first point. The second is-

Anil Chitkara:

Mahesh, if I could just stop you. And I think you've used the term context before, which is a really interesting way to frame that up.

Mahesh Saptharishi:

Absolutely, yeah. So that was going to be the second thing that I was getting at is this notion that historically access control or security at the threshold was a function of identity. Who are you? It depends on who you are. And by the way, who are you could be, "Do I have a ticket to enter this venue or not?" As well as, "I am Mahesh." And the other aspect is, "What am I carrying?" And if I'm carrying something that I'm not supposed to bring into the venue. Or even the opposite, carrying something that I need to bring into the venue. So those are the two elements here. But where access control and access is evolving to, is identity plus context, where it's not just important that I am Mahesh, but it may also be important as to perhaps my history, my background. Am I coming in here with a weapon potentially? And that could be potentially problematic.

Mahesh Saptharishi:

Am I coming here while running a fever? That could be problematic. Am I coming here knowing that I'm COVID positive? That's potentially a problem. Am I coming in here while there's a bolo that is out for me? All of those things are context to me, I identity is verified, I may have every right to enter the facility if this additional context did not exist. But ultimately from a security standpoint, it is something that you want to have that additional context in association with your identity to improve the security posture, but not just the security posture, I think that context also helps to personalize the service that you may get when you enter venue. And you can do this in a very privacy conscious and responsible way because all of us carry phones in our pockets with apps and effectively have this token that we can interact at the threshold with to say, "I'm going to opt into certain personalized services or I'm not going to opt into those personalized services."

Mahesh Saptharishi:

So the threshold now acts as this convergence point for me to communicate my interests to the venue, as I'm going in, including my opt in or opt out to whatever data I would like to share to get those personalized services as well. So, that's I think the other aspect of this. So it's really this identity plus context, which is the evolution of access into a facility, not just necessarily access control, but access with all the services that that access provides us with.

Anil Chitkara:

And I think that we're starting to hear a lot more about that relative to the vaccine. Have you been vaccinated? There's going to be is vaccinated is not vaccinated? Haven't, have not. You can see that at least over the next year and maybe even beyond that. So really interesting. And then the other question is, we were talking about technology and the impact of people, people wear lots of different hats. Something that I've heard you talk a little bit about was this idea of distributing security or distributed security. It's not necessarily just the guard at the door doing that thing. Can you just talk about how you think about that and what you see out there?

Mahesh Saptharishi:

Yeah, so I think as we started off talking about we're entering this event driven world. And in part, we're entering this event driven world, because the days of somebody sitting in front of a security operations center, watching a video wall, hoping that they can detect something that is potentially suspicious or requires attention, I think those days are starting to reduce, especially in the context of COVID now, where you don't want all those people sitting together in the security operation center in the first place, right? Not to mention that tends to be a bit ineffective. But increasingly, that role of a person sitting in a security operation center has been historically part of a certain customer group, who have had a high risk profile associated with them. But even customers with a slightly lower risk profile, that risks may be lower, relatively speaking, but it's still pretty high.

Mahesh Saptharishi:

But those customers don't necessarily have a security operation center, they don't have people sitting in front of a desk, specifically looking to see whether something requires attention and acting as a dispatcher to dispatch people to go investigate or do something. Rather, that role of that dispatcher and the role of the person on the ground is being combined. And these are people who are not sitting behind a desk, rather people who are actually walking around the facility, perhaps doing other things as well, in addition to monitoring the facility as part of their job. And for them, really, the security tasks tends to be interrupt driven, it's event driven, it's this notion that we talked about before where someone taps them on the shoulder and says, "Hey, pay attention to this, there's something here that's important that's potentially happening." And that distributed nature of modern security, physical security stems from that modality, right now, I have a few people now on the ground trying to respond to whatever is happening.

Mahesh Saptharishi:

There's a tap on the shoulder that says pay attention to this. Now there has to be a communication or collaboration between all these people who are distributed along the facility, perhaps with different levels of competence, different levels of skills, whether it is someone who's responsible for the facility, someone who's responsible for security, someone who's responsible for customer service, someone else who's responsible for some other aspect of how some problem needs to be addressed with. And so I think the collaboration between them where in a distributed way, they can share information effectively, look at information effectively, focus their attention and handle that problem. That is the modern distributed approach to security, that trend that we're seeing out there across multiple customers of ours today.

Anil Chitkara:

Right. So an area where I think contexts and distributed security are particularly interesting and close to many of us as schools. I think the characteristics of let's call it the K to 12 schools, leads to or feeds those two trends. What are you seeing in terms of schools and how they're thinking about keeping safe, keeping secure and now [inaudible 00:39:46] COVID and in person and at home and hybrid, which you're dealing with, I'm dealing with, so many parents are dealing with. But how do you think about schools as they reopen more into next year and how they can think about using technology and these trends to make their better, safer, places to be and for our kids to learn?

Mahesh Saptharishi:

Absolutely, I think so. So for one, schools have always been concerned about safety. I think gun violence has been something that been top of mind, increasingly top of mind for a lot of our school districts and school customers of ours. And what COVID has done is it has put a spotlight on the things that the instrumentation that has gone into schools to help with security, but created this new problem now, where it's created bottlenecks, where ingress paths into the school have been something that creates crowds, there's more people standing around. And not only that, I think there's this notion where when people now enter the facility, as students go into the facility, as we're trying to maintain things like social distancing, as we're trying ensure things like people are wearing masks, et cetera and communicating that information appropriately to school administrators, to school resource officers, we get back into this notion that there is this... The context piece of it and also this distributed job function in the schools where the distributor job is really between the school administrators and the school resource officers.

Mahesh Saptharishi:

So when something happens, I think, you want security to be pushed as far out into the perimeter as possible, be able to detect things that you historically were focused on detecting, like guns, like who is this individual? Are there people trying to enter the facility who are not supposed to be in the facility? Is it an authorized visitor to the school that is actually allowed? These are all things that needs to be checked right at the perimeter. So schools often do background checks on visitors as they're coming in, as well. All of this needs to happen right at the perimeter. But once you're within the perimeter, you also want to now understand in the context of COVID, in the context of safety applications, is there a congregation of people that is out of the norm, out of policy? Are we creating new fresh problems with interactions between students without masks, et cetera, that may be a concern? Those events then need to be communicated effectively in real time to the school resource officers, to the school administrators, so they can actually do something about it.

Mahesh Saptharishi:

And ultimately, even in schools, we see this trend towards, if you give the students the right pieces of information through dashboards, through occupancy, give them a nudge to say, "Hey, in this portion of the building, there are way more people here than they really should be." Or, "Hey, you need to maintain more for distance and it's something that's very clearly visual, red or green or yellow in terms of the states of that area. People tend to self-regulate. And I think ultimately, you want to make everybody part of the security and the safety story. And again, that adds another layer of the distributed security responsibility framework. And ultimately, I think what we're seeing collectively here across our customers is how do we now effectively take the feet on the ground and have them be more effective than ever before and empower everybody, the people using the facility and the people responsible for the facility? How do you empower all of them together to improve security and safety in the facility?

Anil Chitkara:

Yeah, interesting. So I do not have a good transition to the next topic. But the next topic, it will be interesting. The other piece I'd like to spend a few minutes with is about the fusion or intersection of physical and cyber. And I think one very recent example and certainly in the news quite a bit is the election, the recent presidential election. I'm curious how you think about physical and cyber relative to the election and the voting that was done and then we can draw that out more into venues, but the election is still top of mind and top of the news for so many. So, we can't go through the hour without talking about it a little bit.

Mahesh Saptharishi:

So I think cyber security and physical security intersect at least one point of intersection is identity. It's like everything that we do when it comes to elections or everything we do when it comes to a person getting access to a facility depends upon verification of whether that person is who they say they are or not. And I think ultimately, a lot of argument about today's elections really stems from the fact that there is no trust that the verification step is happening appropriately. I think ultimately, where we would like to go with both physical and cybersecurity here is the return to trust, where people feel confident that when they say who they are and they enter the facility as that individual who they say they are, there's trust, that is clearly verifiable trust that that is in fact a true statement and they're able to go forward. And I think, with election security today, a lot of good things were already done.

Mahesh Saptharishi:

I think the issue is less about the possibility of there being a cyber security risk, which I don't want to diminish, I think any internet connected system is hackable, the cybersecurity risk persists. But the issue really is an imbalance between that risk and the trust that that risk exists. And I think those two are not appropriately matched. When we talk about the convergence between physical and cyber security here, then a lot of it, I think has to do... Again, starts from that identity problem. When we think about access control, as one example in the realm of physical security, where identity plays a very important part, today, we think about access control as being able to enter a facility or not. That's the access control problem. But as we go into the future, as technologies such as indoor location services, as identity verification at the perimeter, as the digital threshold itself evolves, right now I've fingerprint individuals coming into the facility, access control can turn from macro to micro. Macro being the venue but micro being services within that venue that they can get access to.

Mahesh Saptharishi:

And I think at that level, you're authenticating into that service. And when you authenticate into that service, you are able to consume that service more effectively. This is where I think physical help cyber because today a lot of cybersecurity risks are actually posed as a consequence of social engineering risks, are posed as a consequence of insider threats, where people once they enter a facility, once they enter a building, are able to do things that they really should not be able to do. That is the insider threat aspect of things. And some of it is also a consequence of social engineering. That's where physical security I think helps as we go from the macro to the micro level of access control. I think more broadly, there's plenty of things in physical security that also benefit from cyber security. A lot of the anomaly detection capabilities within the cybersecurity framework, threats in cybersecurity often act as a leading indicator of what may happen within the facility from a physical risk standpoint.

Mahesh Saptharishi:

So I think there's a virtuous connection relationship between what happens in the cyber world, detected threats there and how that can potentially feed into a different sort of readiness, preparation for risk within the physical world and similarly, anomalies and threats that perhaps are observed in the physical world can also feed into potentially what needs to be done in the cyber world to give better context. And we got to come back to this word context again, where again, with that identity, with the nature of that risk, the additional context that is present. Physical where there's a cyber context associated with a cyber where there's a physical context associated with it, helps us more dynamically change and define the nature of the risk that we need to filter, to identify, and then propagate to the right set of humans who do look at it and investigate with the right chunks of data to understand the disposition of the event and also respond to it appropriately.

Anil Chitkara:

Interesting, interesting. This has been such a good conversation, Mahesh, if I think about 50 minutes ago when we started, we were talking about... After the biology and vaccine discussion, getting out of the wet lab. We were talking about these various technologies, AI and big data and analytics and more interestingly, how they apply to people, how they can augment people. And then you sharing how you turn technologies into products. Out shadowing jobs and people and understanding at a very visceral level, what they do and how to improve it and then maybe how to think about the path of technology and the path of their roles and how do we pick a point in the future and develop something that can help that which was very interesting. And then a bit of this discussion on context and distributed security as trends we're seeing out there. And then recently, just talking about cyber and physical coming together.

Anil Chitkara:

So we have a lot of people interested in security and guest services. They're running venues, they are running schools and stadiums. Can you bottom line it for them? So how should they think as they think... Monday morning, how do they think about this conversation in the context of what they're doing first for their venue? And then we'll follow it up and I'm interested in how they should think about it individually. But how do they think about this 50 minutes of rich discussion and what do they do on Monday morning?

Mahesh Saptharishi:

So, that's a really good question. One of the things that I've heard from a lot of our customers, or at least observed, as we've talked is, the design of a physical security approach often is built upon a set of data that tends to be stale, assumptions that are made that are potentially stale, a lot of it is stale data that is gathered from locations, areas that are blind otherwise. That are not instrumented, that are not necessarily something you have direct visibility to, including things like occupancy, including things like flow, et cetera. Monday morning, I think the question you want to ask is, what are my assumptions in the security posture that I've taken so far? What are the assumptions made in terms of the nature of threats that I need to look at? And what is the goodness of that data that led to those assumptions? What are the goodness of the assumptions themselves?

Mahesh Saptharishi:

Are those assumptions blind spots? Where I don't have an opportunity to periodically check those, because those assumptions ultimately end up changing the nature of what needs to be done in terms of the security measures, the risk posture, et cetera, that you need to put in place. So it's really questioning very regularly, and also perhaps on Monday, questioning very regularly, what are my blind spots? And what are the core assumptions that are being made in the design of some sort of process? And are those assumptions valid? And how often do those assumptions potentially change, so that I need to instrument that better, where potentially something that's more of a continuous improvement framework is better than a framework that is based upon static assumptions made for a long period of time?

Anil Chitkara:

It's almost like an agile mentality versus a waterfall mentality in the environment.

Mahesh Saptharishi:

Precisely, yes.

Anil Chitkara:

Good. And then for the individual. So we have a wide array of professionals here that every day are trying to keep their venue, their staff, their visitors safe, or they're trying to improve the services and experience as individual professionals. Maybe Saturday morning, when they wake up, how do they think about continuing to develop their skills, their understanding, their experience, to themselves be tooled for this world that's unfolding in front of us?

Mahesh Saptharishi:

I think a big change that is happening is the democratization of exogenous information. And by that I mean, everything from crime reports to population statistics, to activity statistics across a large geography outside your venue, but perhaps pertinent to your venue. A lot of that is now available readily. And in fact, available via APIs, available via a lot of ways where you can automatically consume that. That can feed into your risk analysis. So I think that the key personal task that I've always tried to spend a little bit of time on, and I can't say that I'm an expert in that area at all, but it's understanding how to now take it into account these new sources of information in better understanding my own level of risk. Risk for the venue, risk for the job that's to be done. How do I now create a framework where I can somewhat algorithmically even look at these data sources and map those data sources and the information it contains into some set of actionable things that are proportional to the risk that the data indicates?

Anil Chitkara:

Interesting. Very helpful to give us things to think about as we think about our own roles going forward. So Mahesh, I want to thank you very much. This has been very interesting, we've covered a lot of ground here. But I think there's a lot of things to think about out of this. Really, even for me as I think about the conversations I have with customers and people out there and then trying to keep people safe and make their experiences more enjoyable, and certainly as economies reopen and businesses reopen, it will happen, it's a little tough right now, but we will get to that point. These are some really interesting things to think about how to use technology to make everything better and make the role of the security professionals better, the visitor experience better across the board. So thank you for your time and more importantly, sharing your insights.

Mahesh Saptharishi:

And thank you for having me. This was a wonderful conversation.

Anil Chitkara:

And for each of you, I'd like to thank you for the work you're doing to keep people safe, to keep their experiences pleasant and frictionless. And hopefully, we're inspiring you to think about new things as you think about your venues and your role going forward. Thank you.

 

 

The Digital Threshold

Enabling Adaptability for Years to Come

Digital transformation is unlocking efficiency and value everywhere as organizations reimagine archaic processes and technology, better equipping themselves with interoperable and flexible capabilities. Within the Digital Threshold vision, venues and facilities can intelligently use data to create a frictionless experience for guests and employees. The result is an entry process that enhances the overall experience instead of diminishing it as it so often does today. Making weapons screening faster and more precise is part of the Digital Threshold vision, but it’s just the beginning.