Episode Transcript
[00:00:02] Speaker A: I'm Alex Stone, former military service member and law enforcement officer, now CEO of Echelon Protected Services, one of the fastest growing private security firms on the west coast. And this is ride along, where our guests and I witness firsthand the issues affecting our community.
I believe our proven method of enacting meaningful change through compassion and understanding is the best way to make our street a safer place and truly achieve security through the community.
[00:00:44] Speaker B: Hi, this is PK Gupta, founder and CEO of Mech Computing. I'm here on the ride along to show how our technology for video analytics is deployed in the field in Portland.
[00:00:58] Speaker C: Hi, this is Jason. No, I'm a computer scientist by training. I'm here on the ride along to understand how the infrastructure deployed to support the first responders to be more effective in the health of the community.
[00:01:13] Speaker D: Hey, it's Alex Dohn.
[00:01:15] Speaker A: Welcome back to the ride along. Today we have two dynamic guests, two tech titans specifically working in the security field and in the smart field, that AI space that's so critical for the future growth of the security industry. Go ahead and Jason. PK, go ahead and introduce yourselves.
[00:01:33] Speaker C: Hi, this is Jason. I'm from South Korea visiting Portland, this wonderful city.
And I'm computer scientist trained and educated here in the United States over at the Stanford University. I was back in Korea for. This is my 10th year working on working with the smart city designers and industrial high tech manufacturing industry to provide the real time process control, process management systems and event generation on the edge side to help them provide the bird's eye view and real time visibility into what's going on in the shop floor. Great, PK, great to be here.
[00:02:12] Speaker E: Sorry. Hi.
[00:02:13] Speaker B: Thanks, Alex. Thanks for this opportunity. So I'm PK Gupta, I'm founder and CEO of Mech Computing. So mech Computing is a company based right here in Portland. We are about six years old. We are focused on using AI, as you were saying, Alex, for security applications. And I'm sure we're going to talk more about that.
[00:02:34] Speaker A: Oh, yeah, for sure. So to get the conversation started, Jason, what you're talking about is your field of expertise is control, how to control smart buildings in a way that makes them safer for everybody.
[00:02:50] Speaker E: Right?
[00:02:51] Speaker C: So in previous generations of competing, integrated, competing with the building management systems, you always waited for event or something happen before you respond.
[00:03:04] Speaker A: Reactive, right?
[00:03:05] Speaker C: Reactive.
[00:03:05] Speaker E: Got you.
[00:03:06] Speaker C: So what we are trying to do is turn the table around, make it proactive.
[00:03:12] Speaker A: I love it.
[00:03:12] Speaker C: And preemptive.
For us to be proactive, we need to collect moral data instead of one data point for say, every second we.
[00:03:25] Speaker E: Do.
[00:03:27] Speaker C: Start listening to the devices. So we classify two different set of data sources. One is human being residents living inside the building. Another is the machines and sensors and controllers.
[00:03:42] Speaker A: Yeah, that makes sense.
[00:03:43] Speaker C: So for human beings, we just wait for them to generate events, meaning push the buttons, yell, or knock on the.
[00:03:54] Speaker A: Door, walk in and out of zones.
[00:03:56] Speaker C: Walk in and out of the zones through the access control systems. But on the other hand, with the machine side, we connect to them in a real time basis. We listen to it, and we also integrate that machine generated, machine sensed information with other sources of data, trying to build intelligence.
[00:04:18] Speaker A: Now, this seems like it would be very difficult because different machines have different code. It's kind of like the.
[00:04:25] Speaker C: They speak different languages.
[00:04:26] Speaker A: Yes, different languages. And the way I talk about this and kind of what you do is I talk about Esperanto. So after World War II, the United nations or NATO, right. We wanted to create a universal language that we were all going to learn, and that language was Esperanto. I've only met two people that know.
[00:04:44] Speaker B: Exactly.
[00:04:46] Speaker A: So the idea is that you've created essentially a tech device that decodes or translates all those different tech languages into a base language, right.
[00:04:58] Speaker C: So, as you just pointed out, to go into a little tech, okay, we call those different set of protocols. Every machine or every sensor has their own set of protocols, but the more protocol or the more devices you place in your premise, the complex.
[00:05:17] Speaker A: It becomes very complex. Exponentially, I would imagine.
[00:05:20] Speaker C: Exponentially.
[00:05:21] Speaker A: We're not just adding. So one plus one isn't two. One plus one is 43.
[00:05:25] Speaker E: Yeah.
[00:05:26] Speaker C: So that's where we come in, and we built sort of set of protocol translators so that we'll be able to understand different set of languages into a single normalized format. That's where the edge computing comes in, and that's where our core competency comes in. We translate all those different set of protocol into a single one so that upstream application or people will be able to understand and make sense of what's going on. And that's what we called contextualization of the edge computing.
[00:06:01] Speaker A: Contextualization of the edge computing.
[00:06:03] Speaker C: So we contextualize it so that you will be able to understand what's really happening.
[00:06:08] Speaker A: So your device is essentially an ambassador, right, in that tech industry, and it's kind of the United nations of devices, right. It's bringing everyone to the table, and it's allowing them to all talk and communicate at the same time.
[00:06:21] Speaker C: Right.
A lot of people thought it would be impossible, but that's impossible. Working on for the last ten years. So we have about 60 different languages sets so that we'll be able to talk to so many different devices and machines.
[00:06:39] Speaker A: And you currently have this edge device program currently running? Actively running right now, yes.
[00:06:45] Speaker C: And we have been focusing on high tech manufacturing. That's where the need was.
[00:06:52] Speaker A: Of course. That makes sense. We have a lot of paper mills in Oregon, and if that paper mill has to shut down for an incident, that's millions and millions of dollars just in like 10 seconds.
[00:07:03] Speaker E: Yeah.
[00:07:03] Speaker C: It's critical to give you an idea of the need. High tech manufacturing. For example, the semiconductor manufacturer out of Samsung. Out of Korea and Samsung. 1 minute downtime means 200 million dollar loss.
[00:07:17] Speaker A: $200 million for how many minutes?
[00:07:19] Speaker C: 1 minute.
[00:07:20] Speaker A: 1 minute.
[00:07:20] Speaker E: Wow.
[00:07:20] Speaker C: So if it's less than 30 seconds, your CEO will be notified right away.
[00:07:24] Speaker A: He might be fired.
[00:07:26] Speaker C: Right. A 32nd window. So that's the sort of the criticality that you're dealing with. But same thing with property management and building management. You can't translate the human laws into the dollar figure. It's higher than that.
[00:07:46] Speaker A: What I think I hear you saying is a part of this interpretive process is being able to interpret human behavior as its own language to then proactively or preemptively stop future actions by that human element, which is obviously going to have more issues than the mechanical and computative element.
[00:08:09] Speaker E: Right.
[00:08:10] Speaker A: They're trying to stop those actions.
[00:08:11] Speaker E: Right.
[00:08:12] Speaker C: So that's the next step. Now that we understand the machines and understand what's going to happen and what's happening and being able to sort of predict what's going to happen in the next, say next minute, next 30 seconds. And we are trying to integrate that intelligence with the video data so that we be able to combine the human data along with the machine data.
[00:08:35] Speaker A: That's where he. And this is where PK comes in. So PK, that predictive element, that is now AI, which I think AI is the new Netscape moment, right.
The netscape was kind of the first browser on the Internet, that kind of that explosive growth under the Clinton administration. So we had this netscape moment with AI where we need that predictive analysis. We're at the point where we can bring all these languages together, explain how Meg is solving that problem.
[00:09:07] Speaker B: Yeah, sure. So to put things in context, what Jason was talking about was all these sensors you find in buildings and factories.
[00:09:15] Speaker E: Right.
[00:09:15] Speaker B: In buildings, especially, access control and building control.
[00:09:19] Speaker E: Right.
[00:09:19] Speaker B: HFC control different languages, and you need to control them.
[00:09:22] Speaker E: Right.
[00:09:23] Speaker B: At mech computing, we are focused on the other source of data, which is cameras, primarily today focused on cameras. And to give you a perspective on why camera input is so important, is there were a billion cameras deployed last year in the world.
[00:09:38] Speaker A: A billion.
[00:09:38] Speaker B: Just think about that one camera for every six or seven people on this planet, right? Most of them are in China, but they're deployed all over the world right now.
[00:09:48] Speaker E: Right?
[00:09:48] Speaker B: And most of the data, there's this data that says that 95% of that data is never analyzed.
[00:09:55] Speaker E: Wow.
[00:09:55] Speaker B: So think about all these institutions, commercial institutions, buildings, factories, everybody has surveillance cameras deployed and all the data is stored away essentially in a video management system. And nobody actually looks at that data to gain insights. So what we are doing at mech computing, and this is part of the intelligent video analytics approach, is you look at the data and as you were saying, proactively try to find insights that are actionable, that people can take action on.
[00:10:29] Speaker E: Right?
[00:10:29] Speaker B: So for an example, for a building, it might mean that you want to monitor, say, the periphery of the building, and you want to monitor people breaking in or loitering.
I think we're going to see that more today. And you can then issue an alert and then take preemptive action.
[00:10:49] Speaker E: Right.
[00:10:51] Speaker B: And you can extend that to other things, other verticals in retail, you can be looking at customers, you can looking at shopping patterns, smart cities. I know we're going to talk about that more. You'll be looking at how to manage a city parking area or a public area more effectively. So there's various applications for video analytics in different spaces.
[00:11:13] Speaker E: Right?
[00:11:13] Speaker B: And we focus on that. Now. One of the key things, Alex, you mentioned earlier is AI, right? So AI is what technology is, what's making this possible today.
[00:11:25] Speaker E: Right?
[00:11:25] Speaker B: So when intelligent video analytics, I should say video analytics, when it started more than 20 years ago, right, when cameras started getting deployed and computer vision came into force, the approach was using computer vision, right, which basically means you take a picture, you look at the pixels frame by frame, and you determine if there's a change in the pixel and you can try to detect motion.
[00:11:47] Speaker E: Right?
[00:11:48] Speaker B: So that's the approach people started taking. And that was like kind of the first or second generation approach to video analytics. Okay, but that, as you might expect, causes a lot of false alarms, right?
[00:11:59] Speaker E: Yeah.
[00:11:59] Speaker A: And I know this to be true, coming from that law enforcement and security background, that a lot of these 911 or BoEC Bureaus of emergency communication centers, they don't even really take these active alarms, even on motion sensors. Correct, because there are so many false alarms.
[00:12:19] Speaker E: Exactly.
[00:12:19] Speaker A: They can't even respond anymore.
[00:12:20] Speaker E: Exactly.
[00:12:21] Speaker B: And so the biggest problem plaguing all these operators of these security systems is false alarms.
[00:12:27] Speaker E: Yeah.
[00:12:28] Speaker B: And I'll give you one data point.
We have this say in auto dealerships.
[00:12:33] Speaker E: Right.
[00:12:33] Speaker B: Or even in buildings.
[00:12:34] Speaker E: Right.
[00:12:35] Speaker B: You have these cameras deployed, you're trying to monitor. So they typically will have special. So besides the video analytics they set up to detect motion or detect people, they will have another system set up in the cloud to filter the false alarms because the false alarms are so prevalent that the control centers or the monitoring stations.
[00:12:54] Speaker E: Right.
[00:12:55] Speaker B: Where they go in. And you got these people, your stereotypical wearing headsets and looking at this picture.
[00:13:00] Speaker A: This is a live monitoring.
[00:13:01] Speaker B: Live monitoring.
[00:13:02] Speaker A: We're calling this a legacy model.
[00:13:04] Speaker B: Yeah, this is legacy model. Live monitoring. Whether you deploy all these people, expensive resources and all they do, and it's a very boring job and they find terrible to keep these people employed.
[00:13:14] Speaker E: Right.
[00:13:14] Speaker B: It's because you're just looking every frame by frame and trying to determine.
[00:13:18] Speaker A: And they're responsible for thousands of cameras because like you said, we just deployed another billion last year.
[00:13:24] Speaker E: Right.
[00:13:25] Speaker A: Thousands of cameras are being. And then they have to choose from those thousands and thousands upon when they're alerted.
[00:13:33] Speaker E: Correct.
[00:13:34] Speaker A: And they have maybe 10 seconds to determine if they think it's a criminal activity.
[00:13:38] Speaker B: And the time, typical time to respond for these legacy systems, right. Is about ten minutes or plus because the event happens, it goes in a queue.
[00:13:48] Speaker A: That's right.
[00:13:49] Speaker B: Then you filter it. There's another queue. Then you go to the monitoring station. Monitoring station staff might be short staff. So you'll go into a queue in the monitoring station and then the monitoring station. Somebody's going to get to it, look at it, decide. Oh, yeah. This is a real event happening, real break in. So I need to send somebody, then they contact the guard. Now typically what happens in a typical scenario, you contact a security company who then contacts another security company who then deploys the guard.
[00:14:14] Speaker E: Yes.
[00:14:14] Speaker A: Dispatch to dispatch to dispatch.
[00:14:16] Speaker E: Dispatch. Right.
[00:14:16] Speaker B: So if you look at the whole chain, it takes you like ten to 15 minutes before somebody shows up.
[00:14:23] Speaker A: This is a real problem in our industry.
[00:14:25] Speaker E: Exactly.
[00:14:25] Speaker A: As you're saying.
In Portland, Oregon, the average response time for law enforcement to a critical incident is about 1415 minutes.
[00:14:36] Speaker E: Yes.
[00:14:36] Speaker B: So that lines up with the national average. Also. It is some places less typically, some.
[00:14:43] Speaker A: Places are around seven to eight minutes. But the main problem is the average residential, commercial burglary or assault is less than seven minutes.
[00:14:51] Speaker B: Yes.
[00:14:52] Speaker A: So what we're talking about is reducing, using AI and tech, knowing when someone access to building, verifying that through AI and then getting a direct action response in less than seven minutes, reducing that.
[00:15:05] Speaker B: Less than a minute.
[00:15:06] Speaker E: Right.
[00:15:06] Speaker B: The ideal is now you use AI to determine with high accuracy, because what AI allows you to do is cut down on those false positives dramatically.
[00:15:17] Speaker E: Right.
[00:15:18] Speaker B: So we've seen evidence where we've cut down false positives from hundreds at night across, say, ten typical ten cameras in a location to maybe five handful. So you cut down the false positive. So now the alerts coming in, you can trust them.
[00:15:33] Speaker E: Right.
[00:15:33] Speaker B: So now that you trust the alerts, you don't have to filter them, you put them in a queue.
[00:15:37] Speaker E: That's right.
[00:15:38] Speaker B: You can send them directly to the guard who's stationed somewhere.
[00:15:41] Speaker E: Right.
[00:15:42] Speaker B: And the guard could be having security person, could be having a mobile app.
[00:15:45] Speaker E: Right.
[00:15:46] Speaker B: You can send the alert straight to the app and they can take action.
[00:15:48] Speaker E: Exactly.
[00:15:48] Speaker A: And this is what we developed was this front end piece of this mobile app, right?
[00:15:53] Speaker C: Correct, exactly. And that video information would be reinforced or validated with all the sensory information.
[00:16:00] Speaker E: Yes.
[00:16:01] Speaker A: Like fire sensor information, humidity access control, access control, echolocation, window brake information, even possible thermal tracking of thermal tracking and.
[00:16:14] Speaker C: Air displacement, radar, lidar.
[00:16:17] Speaker A: All of this coming into one central location and then AI reading everything.
[00:16:21] Speaker B: Contextual analytics.
[00:16:22] Speaker E: Right.
[00:16:22] Speaker B: So combining video, that's the next step.
[00:16:25] Speaker E: Right?
[00:16:25] Speaker B: Combining video and sensors together. So not just looking at, relying on video itself. If you rely on, with AI, you get very good results. But now, next step, you combine the.
[00:16:34] Speaker C: Video with physical things, information, and the.
[00:16:37] Speaker B: Information is even more accurate.
[00:16:39] Speaker E: Right.
[00:16:39] Speaker B: And so you can now do what's called contextual analytics or contextual analysis, where you can have rules. Now that if my sensor, like, say you have a door access control system, right? And you can say, if my access control system is telling me there's an alarm and my video is also showing me there's a person, then I'm like 100% sure that there's a break in happen, and then I can rely on that information with much more accuracy. So that's the best combination where you.
[00:17:08] Speaker C: Combine or if the video analytics shows that there's something event, we also rely on machine data to validate it. And we also have, based off of the machine data, we'll be able to tilt and zoom the video camera in such a way that you would be looking at right where the automatic is aligned.
[00:17:27] Speaker A: Automatic. That's exactly what AI driven access control. AI driven control systems.
[00:17:34] Speaker C: So that he works with the dispatcher, works with the first responders to give them the area of focus. So that instead of wasting their time. That point, first floor, second window, that's where his focus should be.
[00:17:50] Speaker A: And then they can track not just within different analytic zones, within one camera stream, but then also camera to camera.
So you get a unique tracking id, whether through facial recognition, gate height, weight, these things. So that tracking id, you can track this person all the way across the camera.
[00:18:10] Speaker B: You can follow the person, see what they're up to. If they're going into a zone they're not supposed to go into, you can flag that.
[00:18:16] Speaker E: Right?
[00:18:16] Speaker B: So you can track people without using facial recognition. Using AI based techniques, you can track people with unique ids. So that's one of the benefits.
[00:18:24] Speaker A: So I'm going to mention a case study here and I'm going to not mention the names.
I was approached by a client. This individual represents an ownership group that owns properties in multiple states, high end commercial properties. They had an incident at a property where a domestic occurred between a top executive.
The spouse came the next day, was able to access the garage because they had their spouse's vehicle, was able to access the building because the spouse had a second badge that this person had access to. And then they were able to get to the top floor. And during a board meeting, they actually killed their spouse in front of their entire team.
[00:19:12] Speaker B: Wow.
[00:19:13] Speaker A: And so at what point will we be at a level of intelligence, actionable intelligence, where we can deploy things like stingers? So in law enforcement, we have these stingers for the guest that's like a cell tower. The CIA does it when people turn on their all, we all do this, not just the CIA, but we're tracking cell phone numbers and cell phone data and these ip addresses, these unique addresses on these tech devices. At what point could we have, if their Internet would have been open, said, hey, we know that IP address, we know that license plate. We have facial recognition. Right.
Could AI then initiate a lockdown by floor or by building or by area?
[00:20:02] Speaker E: Yeah.
[00:20:03] Speaker A: How far away are we from this?
[00:20:05] Speaker B: I would say what you're bringing up is what in the industry, the big trend is the convergence of physical security and cybersecurity.
[00:20:14] Speaker E: Right.
[00:20:14] Speaker B: Because what's been happening is people typically talk about physical, have been talking about physical security, but cybersecurity is now getting more and more importance.
[00:20:23] Speaker E: Right.
[00:20:24] Speaker B: And the link is in the industry they're seeing is they're both kind of relying on each other. Most of the, by the way, cybersecurity break ins happen because of physical security flaws.
[00:20:33] Speaker E: Right? Yeah.
[00:20:33] Speaker A: That's very critical to expound upon that because, like, you're saying cybersecurity, which is your networks within a building or within a network.
[00:20:43] Speaker E: Right.
[00:20:43] Speaker A: Most of those breaches happen because of physical security issues.
[00:20:46] Speaker B: For example, the latest issue, right, where this person took all these pictures of these sensitive documents and posted them on this channel. Now, he had physical access to those things.
[00:20:57] Speaker A: Exactly.
[00:20:58] Speaker B: Or even if you look at Snowden.
[00:20:59] Speaker A: Sometime back, Edward Snowden, he was able.
[00:21:03] Speaker B: To get access to all the data because he had physical access to the building, to the site.
[00:21:06] Speaker E: Right.
[00:21:07] Speaker B: So it's always a combination of physical security and cybersecurity. Now, in this case, I'm not very familiar with the details, but the trend is if you are able to define the environment where on a single platform, if you can not only analyze physical security threats, like somebody trying to break in or you're monitoring somebody, and then you see some cybersecurity patterns in the network, like somebody trying to access the network that not. They're supposed to.
[00:21:33] Speaker E: Right.
[00:21:33] Speaker B: Who know who's not supposed to. Then you can try to combine the two and then define an event and prevent incidents like that.
[00:21:39] Speaker A: That's amazing.
[00:21:39] Speaker C: That's what our industry calls endpoint analytics.
[00:21:43] Speaker A: Endpoint analytics, yeah.
[00:21:44] Speaker C: So you have all the circumstantial data, but that data would be married into what he or she or that particular cell phone is doing. Is it in the regular pattern? Are we seeing something different today? Is she doing something anomalous today? Then? Triggers a lot. Automatic triggers, alarm. And another layer of analytics is run on just to ensure that she or he is doing everything typical, or that particular device is doing something natural, or is it something anomalous, something different? Do we need to pay our attention onto that particular device or a particular person of interest? That's what we call endpoint analytics. That's something that is being added into the physical security and cybersecurity these days.
[00:22:38] Speaker E: Yes.
[00:22:38] Speaker B: Anomaly. Any kind of anomaly detection, right.
[00:22:40] Speaker A: Could be of human behavior.
[00:22:42] Speaker B: Human behavior or network behavior.
[00:22:44] Speaker A: Oh, really?
[00:22:45] Speaker E: Yeah.
[00:22:45] Speaker A: Network behavior, too.
[00:22:46] Speaker B: Networks typically will have a pattern of behavior, right. You'll see some attempts coming from certain IP addresses, right. And suddenly one day, you see you're getting a lot of requests for some other IP addresses, right. Which could denote some kind of attack happening. So, yeah, the behavior patterns you can set up for both people and networks.
[00:23:06] Speaker E: Right.
[00:23:07] Speaker B: And then monitor that. And then that's the benefit of monitoring both of them together and then seeing turner patterns between them.
[00:23:13] Speaker A: So when we're discussing the convergence of AI and controls, AI driven controls.
Let's think of a case study and with the idea of understanding the time frame that it would take, subsecond. Sub millisecond. So we have ten sensors in this room. All these sensors are triggering anomalies. Right. The AI is taught to say when these anomalies occur in this pattern, it's likely going to be this type of incident. How long is all that taking?
[00:23:48] Speaker C: Well, to give you an idea, let me take an example of a quite recent incident. I can't divert where it happened, but this is a very critical facility, and the fire detection is critical because if you could detect the fire and put it out, say, within 3 seconds, you're talking about 100 k's of losses.
If it lasts more than 10 seconds and start generating dust and particles and all that, that's hundreds of millions developed a loss. So you have to put out the fire within 10 seconds.
[00:24:25] Speaker A: So that seven second gap is the difference between 250,100 multimillion.
[00:24:30] Speaker E: Yes. Wow.
[00:24:31] Speaker C: Okay.
[00:24:32] Speaker A: Seconds.
[00:24:33] Speaker C: So another thing for us to take into the consideration is whether this is accurate information, actionable information that we could really act on, because in this particular facility, you don't put the water to put out the fire. You put the co2 into the air so that you suck up all the oxygen.
[00:24:53] Speaker E: Got you.
[00:24:54] Speaker C: That's how they put out the fire. But if there's anybody, any human being, any worker in the facility, he lung, would collapse instantly.
So the accuracy of the information is critical, the response time is critical.
We deployed five, six different type of sensors, and all the five different sensors told me that this is a critical, infamous critical. The fire. Fire is about to start. And we have video analytics integrated and access control integrated. So we know for sure that in this particular zone, there's no workers and video analyts tell me that for the past seven minutes, there's no excess control, no human being coming into the zone.
[00:25:39] Speaker E: Wow.
[00:25:40] Speaker C: And we have a one video shot around it. Critical, no human. Then we just put out the fire by such an way of instant extinguisher. That's how it happens.
[00:25:56] Speaker A: How many accolades? How many accolades did you get?
[00:25:59] Speaker C: I get two plaques of recommendation and appreciation from the client.
[00:26:04] Speaker A: That's impressive. If I owned that company, I would keep you very close to me, for sure.
Let's kind of bring it back to Portland.
Tech is important, and saving microchips are important, but I think what's going to really drive the industry is saving.
So here in Portland, there's a lot of people that are pushing for echolocation. It's called shotspotter.
[00:26:30] Speaker E: Right.
[00:26:30] Speaker A: They want to be able to identify where firearms have been discharged in order to allocate resources, real time resources, law enforcement in the field.
As I understand it, AI is not really driving this. But could you kind of give a scenario where if AI was interpreting a lot of this data, right, maybe cars running red lights, these types of things, would that make it more effective?
What are we really talking about when we're talking about securing areas and making people safe?
[00:27:06] Speaker E: Yes.
[00:27:07] Speaker B: I make a general statement.
[00:27:08] Speaker E: Right.
[00:27:08] Speaker B: AI, in general can help with any problem like this, right. It's how you apply AI carefully.
[00:27:14] Speaker E: Right.
[00:27:15] Speaker B: For example, you have to be sensitive to the fact even in physical security surveillance, you don't want to do facial.
[00:27:21] Speaker A: Recognition, for example, which is illegal in.
[00:27:23] Speaker B: Portland, which is illegal many places, including Portland. And so you can do without it, right. You can use techniques without facial recognition and still get your job done.
[00:27:31] Speaker E: Right.
[00:27:31] Speaker B: Similarly for firearm detection. So there's various techniques, right. One of the techniques is people use acoustic sensors or special kind of sensors or even vibration sensors, so that if they detect something within some distance, they know there's a gunshot. Gunshot has a very specific kind of vibration, so they can detect that. And that will give you, say, one indication, right. Something has happened. You can couple that with other sensors, right. So you can have an acoustic sensor, a vibration sensor, a video sensor, like a video image. So video image. Now, people are deploying this in various schools now, right.
There was one latest one I saw in Philadelphia, I believe. School system.
[00:28:11] Speaker A: Yeah, Pittsburgh or Philadelphia. It was definitely Pennsylvania.
[00:28:13] Speaker B: And somewhere in Pennsylvania, right, where they're deploying this. And the idea there is that if somebody's entering a school with a gun, they've got these cameras that can detect potentially and issue an alert before the incident happens.
[00:28:26] Speaker A: So a firearm verification.
[00:28:32] Speaker B: System.
Well, the promise is, right, that, of.
[00:28:37] Speaker A: Course, no one's walking in.
[00:28:38] Speaker B: No one's walking in, brandishing a firearm, right. But the promise is that, okay, the camera is going to detect a firearm and issue an alert.
[00:28:47] Speaker E: Right.
[00:28:47] Speaker B: Typically, the firearm is going to be concealed.
[00:28:49] Speaker E: Right? Typically. Right.
[00:28:50] Speaker A: And there is the possibility of indexing. Indexing in my field is the subconscious behavior, behavior of constantly needing to touch something that's illegal, like a firearm or drugs.
I think it's three times every 10 seconds.
[00:29:08] Speaker E: Right.
[00:29:08] Speaker A: If someone.
[00:29:09] Speaker B: You can look for those behaviors, you can look for those behaviors with video analytics and trying to reinforce that something is suspect. So behavioral analysis can be done very effectively with video analysis and video analytics and AI. But I think the promise of detecting the firearm physics itself, it might be over promised. So you got to use some other techniques, obviously, metal detectors help, but is it practical? So it has to be a combination of things, right, to make the problem, to solve the problem.
[00:29:36] Speaker A: So let's stay on the schools area, because that's something that's near and dear to me. When I was in law enforcement, I worked a lot in the schools, right. And I also did some juvenile sex crimes, things like this. So at what point is there a point in the future, in the near future, where because a handgun is hard to. Is easy to hide, and usually if a student's bringing a gun to school, it's almost always a handgun. But most school shootings are long guns. So if someone exits their vehicle with a long gun and they're walking towards an entrance of a school, at what point in the future will AI be able to detect that and then engage an AI driven access control system to then start locking the school down or go through an app like ours, Sitrep, where then an administrator.
[00:30:29] Speaker E: Yeah.
[00:30:29] Speaker A: And then they can then engage the access control on their phone.
[00:30:32] Speaker E: Right.
[00:30:32] Speaker A: At what point are we, if the.
[00:30:34] Speaker B: Person is walking with the gun, carrying the gun, then you can very accurately detect it.
[00:30:39] Speaker E: With AI.
[00:30:40] Speaker B: AI models can be trained to detect different kinds of guns.
[00:30:43] Speaker E: Right.
[00:30:44] Speaker B: The long guns. But if it's a concealed, then you cannot. But long guns.
[00:30:47] Speaker E: Yeah.
[00:30:47] Speaker B: You can train it very effectively if the person is walking with that and then issue an alert to a local law enforcement or somebody to try to clamp down the place.
[00:30:56] Speaker A: And then not only could AI then lock the building down, but it could then give real time information to law enforcement as to where the individual is within the building.
[00:31:07] Speaker E: Yes.
[00:31:07] Speaker A: And this is track the person, someone who's been in critical incidents, someone who's been in.
I've been there and I've done that.
[00:31:14] Speaker E: Right.
[00:31:15] Speaker A: Finding the person that's engaging in the criminal activity usually is what takes the longest point of amount of time and having that situation report coming through in real time, in subsecond, sub millisecond time, that's what's going to be able to allow that direct action or that direct asset to then eliminate that threat.
[00:31:36] Speaker C: I like to draw the attention to the fact that it requires infrastructures for the video analytics to be more effective. It would require higher resolution cameras. Not the. That's not going to work, but higher resolution.
Higher resolution means you need to take more video into the network. So most of the schools and facilities that I visited, their infrastructure is not really ready. That's one of the reason that the modern technology doesn't really deliver for the modern technology to really deliver not enough information structures.
[00:32:15] Speaker E: Wow.
[00:32:15] Speaker C: Better networks, better cctv cameras and high resolution sensors to deploy those, you need some government interest, public interest, into improving the infrastructure. That's what really counts.
[00:32:31] Speaker A: So investment in the public infrastructure at the school level to get the right type of infrastructure in place that can deliver enough ones and zeros to make it effective and quick.
[00:32:41] Speaker C: I mean, interesting thing is, in Asia, where I'm come from, the gun is not really much of an issue, of course, but when I was doing a smart city project, consulting project in the San Diego, the gun detection was such a problem. But how do you help first responders to locate and come to the right spot where the gun is discharged? You could detect the gun and give you the 200 meters radius of this, I think, is the area where the gun is detected. But if you're going to the downtown city. 200 meters radius. That's a lot of buildings.
[00:33:20] Speaker A: Yeah, that's a lot.
[00:33:21] Speaker C: One of the idea that we did was, let's put these smart led lights with the control systems. We will be able to locate the work with the gunshot detectors and locate the exact locations within, say, two blocks and have the lights turn blinking to give you a visual cue for the first responders. This is where you want to be.
[00:33:48] Speaker E: Wow.
[00:33:48] Speaker C: That's one of the idea that we.
[00:33:50] Speaker A: That's an amazing idea.
[00:33:51] Speaker B: No, that's actually very potent. We are working on a project, smart city project, which is relevant to managing cities where they have these special lamps.
[00:34:01] Speaker E: Right.
[00:34:01] Speaker B: They deploy in the street. The lamps obviously provide light, but besides that, they have cameras in the lab and they have 5g modems in the lab.
[00:34:10] Speaker E: Yeah.
[00:34:11] Speaker B: So they basically become hotspots.
[00:34:13] Speaker E: Right.
[00:34:13] Speaker B: So you don't have to use Wifi, they're hotspots. And the camera analytics is basically trained to look at people. And if anybody's in distress, any person, for example, say somebody has fallen or somebody's running in distress and asking for help, the video analytics sees that and generates an alert right away to get help and the lights start flashing.
[00:34:33] Speaker A: Well, that's great.
[00:34:34] Speaker B: So you know where it is.
[00:34:35] Speaker A: So you have a location.
[00:34:36] Speaker B: This technology is getting deployed now.
[00:34:38] Speaker A: That's amazing. Yeah, this is really amazing.
[00:34:40] Speaker B: It is available, extending from buildings to smart cities.
[00:34:43] Speaker E: Right.
[00:34:43] Speaker B: Is, I think, what we want to see today also.
[00:34:45] Speaker A: So I'm going to pivot because there's always the question of Skynet. Right. And for Skynet is a reference to a movie where the machines take over and they eventually determine that they need to create an apocalypse to destroy the human race for self preservation. And you have that and then you also have the general.
I would say there's a general anxiety just about the idea of a surveillance state. And I think a lot of this comes from just the development of control countries. We'll just say control countries. These are countries that we feel don't play well with the general population of the other countries, right, that are represented like United nations. And these control countries tend to have more of surveillance states.
They're using a lot of this technology really to figure out what their individual citizens are doing to give them more control over their population. And so I think on the entire political sphere, everyone is concerned about this.
How do we make sure that what we're doing doesn't become that? But let me preface it with this.
We're really lucky because in America we have a constitution. So in London there are larger surveillance know. I know that a lot of times New Zealand has chosen to do case studies because they don't have the issues that you would have in America to monitor things on a city level in public view. So with all that being said, kind of give me an idea of how we continue to deploy this technology, but to do it in a way that has oversight.
[00:36:27] Speaker E: Yeah.
[00:36:27] Speaker B: So any technology, and this is cliche, but any technology can be used for.
[00:36:32] Speaker E: Good or bad, right?
[00:36:33] Speaker B: So we know that. So same thing goes for AI and video and everything, right? So some of the states you talked about, these controlled states, they're using the technology to monitor the inhabitants, the citizens. Right. And trying to get better control of them. But the same technology is used in, let's say, liberal states or western states.
[00:36:53] Speaker E: Right.
[00:36:54] Speaker B: For, say, law enforcement in London, which has a very extensive network of cameras everywhere. By the way, New York has more cameras than London. People don't know that.
[00:37:02] Speaker A: Oh, I did not know that, yes. Really?
[00:37:04] Speaker E: Yes.
[00:37:05] Speaker B: But you don't hear about that much.
[00:37:06] Speaker A: And fewer residents.
[00:37:07] Speaker B: Yeah, but it's all there.
[00:37:08] Speaker E: But you don't hear about it.
[00:37:09] Speaker B: London, you hear about it because they talk about it all the time and they use it effectively for law enforcement.
[00:37:15] Speaker E: Yeah, they do.
[00:37:16] Speaker B: People, incidents happen. They're able to track the person down.
[00:37:19] Speaker A: Going into the well, and their mass transit system almost requires it underground, like the bombing of the tube or subway, all these trains have to be rerouted and stopped.
[00:37:31] Speaker B: So it all comes down to technology regulation, right? And the same thing goes for AI. If you recall, as technologies were rolled out, people always, even before AI today, right? If you go back, say, 1020 years, when the Internet technology first came out, people were like, this is going to take over the world and destroy our kids and everything's going to happen. And even before that, I was looking at a report. You'll find a surprise when calculators were first introduced back in the 50s for children in school.
[00:37:58] Speaker A: So going from the abacus to a calculator.
[00:38:00] Speaker E: Right.
[00:38:01] Speaker A: Okay.
[00:38:01] Speaker B: The teachers went on strike in american schools because they thought that the children are not going to be able to learn math anymore.
[00:38:09] Speaker E: Wow. Right.
[00:38:09] Speaker B: Because they're going to use calculator.
[00:38:10] Speaker E: Yeah. Right.
[00:38:11] Speaker B: So we went that wave, then we went through the Internet. Now AI, now, you might know this. This is happening in Congress. Now, this is not the best place to make regulations on AI, but Congress is trying very hard to come up with. And the AI leading guys from the leading companies have themselves told Congress that please regulate us.
[00:38:33] Speaker A: Really? Yeah.
[00:38:34] Speaker B: So the Google, Facebook, I mean, meta, Microsoft, all of them have appealed to Congress that please introduce some legislation now. Because with the latest AI technology, if it is not controlled properly, people can use it for mischievous purposes.
[00:38:53] Speaker E: Right? Yeah.
[00:38:53] Speaker B: So there is a demand right now from the industry itself to regulate it.
[00:38:59] Speaker E: Right.
[00:38:59] Speaker B: And I think the government has to step in globally.
[00:39:02] Speaker E: Right.
[00:39:02] Speaker B: And Europe is trying to.
[00:39:03] Speaker A: Is this to protect the industry from future lawsuits, likely, or not just civil.
[00:39:07] Speaker B: Lawsuits, but actually, no, these people are actually concerned that somebody can misuse the technology and cause the threat. You talked about Skynet, they are talking about threats like that right now.
[00:39:19] Speaker E: Yeah.
[00:39:19] Speaker B: It's becoming more and more real. The AI is becoming so powerful now that that is a possibility.
[00:39:26] Speaker E: Right.
[00:39:27] Speaker B: So they want to regulate that right now to prevent anything like that even remotely happening. So with the proper regulation, we cannot stop it.
[00:39:36] Speaker E: Right.
[00:39:36] Speaker B: So we have to regulate it and use it responsibly for the benefits.
[00:39:40] Speaker E: Right.
[00:39:41] Speaker B: And manage it that way.
[00:39:43] Speaker C: So inclusive technology, fair technology, ethical technology, that's really a critical concept. Up until now, if you look at the technology, it's been the feature development, development was the focus.
I think now we are moving on, matured enough technology, mature enough and deployed enough. Now we are moving into the next phase of how could we more inclusive, how could we design the system or technology in such a way that it is fair, ethical? So we are moving into that phase.
[00:40:17] Speaker A: So we're going to a stage of discipline where we have to discipline ourselves to do good regulation framework.
[00:40:25] Speaker E: Right.
[00:40:25] Speaker C: So regulation is one way and another was internal, intrinsic self control, self control by designers and the scientists to put in those features so that you would be inclusive, ethical and accurate and reliable.
[00:40:43] Speaker A: My concern, this is just my concern. And it just kind of happened. But when you have control states whose main force in the world is theft, there's like a kleptocracy almost of you have governments that are kleptocracies. They are ruled by thieves because they're driven by the greed of their own profits and to protect those profits through one party rule. Right. And so when you have that main desire, I would just call it greed.
And we're not talking about economic systems, we're talking about the individuals in control. Right. When greed is the main purpose for the theft and reuse and repurpose of this, is it possible to have that discipline? Because at some point these control states, because of their desire, they can abuse the technology. They can abuse it and then also learn faster.
[00:41:46] Speaker E: Yes.
[00:41:47] Speaker B: And they are doing that. They're ramping up and there's potential for abuse. Definitely. I mean if the governments themselves sanction it. Right, or use it for their own purpose, they can abuse it.
[00:41:57] Speaker E: Right.
[00:41:58] Speaker B: And even today's technology, they're abusing it and tomorrow's technology, they'll continue to abuse it. But the safety is that on the flip side, we have to be prepared to be able to protect ourselves against attacks from.
[00:42:10] Speaker A: Yes, because this is AI is moving. It's all the convergence of cyber AI, direct action, physical security. The convergence is real and we do need to protect ourselves.
[00:42:21] Speaker B: We have to protect ourselves with proper.
[00:42:23] Speaker A: Regulations, with proper behavior and superior ones.
[00:42:28] Speaker B: And zeros and superior technology. We have to stay ahead of the curve.
[00:42:32] Speaker E: Right.
[00:42:32] Speaker B: That's the only way to protect ourselves.
[00:42:34] Speaker E: Yeah.
[00:42:34] Speaker A: And the only way to protect ourselves against future intellectual property theft is this technology.
[00:42:39] Speaker B: Yes.
[00:42:40] Speaker E: Right.
[00:42:40] Speaker A: This technology will be the buffer from control states stealing our tech in the future.
[00:42:46] Speaker B: It'll help us protect barriers in national boundaries, everything, both cyber and physical.
[00:42:52] Speaker A: Okay, so we could talk about this forever.
[00:42:55] Speaker E: Forever.
[00:42:55] Speaker A: But this is a ride along, right? We're going to go on a ride along. We have some of this technology deployed here in properties and we're going to go, we're going to test it.
The security company that's responding, I'm currently CEO of. They're not going to be notified of when we're testing the tech, but we're going to test these analytics and we're going to test the response time.
[00:43:16] Speaker C: Okay, cool.
[00:43:17] Speaker A: And then after that we're actually going to meet with a developer, a local developer that's creating a large community and they want that whole community to be smart. So we're going to run through that and we're going to look at what it takes to develop a smart community. Okay, sounds good.
[00:43:32] Speaker E: Great.
[00:43:33] Speaker A: All right, let's write good.
[00:43:34] Speaker E: Okay.
[00:43:35] Speaker A: We're meeting with Adam Schneider. He actually just pulled up. He's the CEO of Piston Overwalk Services, which is a value added distributor of Meg. So they're deploying with other companies this Meg bass solutions in the field.
[00:43:49] Speaker E: Hey, Adam.
[00:43:50] Speaker A: How's it going, man?
[00:43:50] Speaker E: Hey, Adam. Good to see you, bro. Good seeing you, too. Hey, Adam.
[00:43:53] Speaker C: Hey, Jason.
[00:43:54] Speaker B: Jason, nice to meet you, Adam. Nice to meet you.
[00:43:56] Speaker E: PK. Yeah.
[00:43:57] Speaker A: You ready to go?
[00:43:58] Speaker E: I am.
[00:43:58] Speaker A: You got a couple sites for us picked out?
[00:44:00] Speaker E: Yes, I do.
[00:44:01] Speaker A: All right, let's hit it.
[00:44:02] Speaker D: Let's go.
[00:44:02] Speaker E: All right, great.
[00:44:08] Speaker A: So this is the site?
[00:44:10] Speaker E: Yeah.
[00:44:10] Speaker D: So, prior to us launching on this site, there was a lot of drug use, loitering.
It was quite a lot to handle for the clients and customers coming to visit. So, once we deployed here, we've seen a significant decrease in crime activity, mostly because of the response time. So when someone walks in there, we're going to track the notification being triggered and we're going to see how long it's taking the security guard company to respond.
[00:44:38] Speaker A: So this is their analytic is in this area?
[00:44:41] Speaker E: Yes.
[00:44:41] Speaker D: There's a camera that's back hidden in that cubby way.
[00:44:44] Speaker A: Okay, I see top right side.
[00:44:45] Speaker E: Yep. Exactly.
[00:44:46] Speaker D: And this is the zone that we've drawn so we can.
[00:44:49] Speaker C: These are the virtual fences. Exactly.
[00:44:51] Speaker D: So this is the virtual fence. So we know.
[00:44:53] Speaker A: And this is the meg Uiux.
[00:44:55] Speaker D: Yes, correct. So this is the Meg video analytics solutions that's powering our company. And so when someone walks into this zone, the threshold is set for 15 seconds. So after 15 seconds, you're going to see the notification turn red, and it's going to alert the officer.
[00:45:10] Speaker A: That's working. Do you mind if we just walk someone in there so he can kind of.
[00:45:13] Speaker D: Please do.
[00:45:13] Speaker E: Yeah.
[00:45:14] Speaker D: Let's go ahead and test their response.
[00:45:16] Speaker E: Wow.
[00:45:17] Speaker A: We won't stay in there more than 15 seconds.
So you can see us on the iPad.
[00:45:22] Speaker E: Yes.
[00:45:23] Speaker D: So keep going in.
[00:45:24] Speaker A: Go in.
[00:45:26] Speaker D: Keep going and just hang out there for. Stay there.
[00:45:30] Speaker C: The owner of the parking garage are having trouble watching these.
[00:45:37] Speaker A: We don't really look like criminals, but I'm trying. I tried my best.
[00:45:42] Speaker B: So Adam did it.
[00:45:46] Speaker D: Notification went for the officer, and now we're waiting for officer response.
[00:45:51] Speaker B: So we stay here or we come out.
[00:45:52] Speaker D: You guys can walk out.
[00:45:53] Speaker B: Stay there.
[00:45:57] Speaker C: Do you also provide the audio alarms?
[00:46:01] Speaker A: Some clients do like to have the audio. The problem is the audio real interdiction requires us to identify the suspect. So if you do an audio, you're kind of telling them you're not coming, number one. Number two, if they leave and you don't identify them, they'll just come back later.
[00:46:18] Speaker E: Right.
[00:46:18] Speaker A: So what we want to do is we want to provide the opportunity for the security company, that real direct action asset, to get in the field, to beat the cycle time, capture the person in the activity, identify them and trespass them, because ultimately that trespass is what's going to stop them from coming back.
[00:46:38] Speaker E: In the future.
[00:46:40] Speaker A: You'll get maybe 20% effective results with a call down, but you'll get an 85% to 90% effective results for drop in incident.
[00:46:49] Speaker C: The reason I was asking about the audio warning is that in Korea you have to keep the warning first. That's a part of the regulation.
So the video will have to be, once it is triggered, you have to have an audio playing, you are trespassing. It can be you have 10 seconds.
[00:47:06] Speaker A: To leave or AI generated. It has to be the real person.
[00:47:10] Speaker C: Yeah, it's dependent on the scenario recorded audio being replaced.
[00:47:14] Speaker A: That's what we prefer. We would prefer a general audio recording rather than having to pay for that.
[00:47:18] Speaker D: Physical asset and some of our properties. We've been able to API the existing technology for a voice down and so now we've gotten an automated voice down as well as this notification trigger. So real time officers aware as well as a voice down being deployed. In some instances there's lights that are attached to that. Lights as well as a voice flashing.
[00:47:37] Speaker C: And voice, but audio.
[00:47:39] Speaker D: The biggest difference is that while that's taking place, an officer is already in route whereas in typical legacy models that's taking place. But nobody's been notified for response yet.
[00:47:49] Speaker C: So you'd be able to check off all the checkboxes if needed to file charges.
[00:47:55] Speaker E: Exactly.
[00:47:56] Speaker A: Yeah, exactly.
[00:47:57] Speaker D: And all the information stays in the arresting party's hand.
[00:48:00] Speaker E: Absolutely.
[00:48:00] Speaker D: So the officer can use that information, they can testify, we can compile all.
[00:48:05] Speaker A: Of that data and they'll be able to identify the individual with the video? Sure, they'll be able to identify them.
[00:48:12] Speaker D: So we're at two minutes and 40 seconds.
[00:48:15] Speaker B: So do you know if the alert has gone to the officer? Officer, yes, it has. So do you know if the officer has received it and is responding right now?
[00:48:22] Speaker D: Within 1 second the officer has acknowledging. Yes, exactly. They get the notification.
[00:48:26] Speaker A: So within 1 second of the analytic having been triggered, the guard who's already patrolling the area was notified on like a device, correct?
[00:48:35] Speaker D: Their cell phone?
[00:48:36] Speaker A: Yes.
[00:48:37] Speaker D: A notification goes directly to their handheld with a link, so they can click the link. And the link is going to look similar to an event history. So this is what will pop up.
And so the guard is going to see, hey, there's three individuals that are.
And so the officer will see this and they'll be able to know what they're responding to and what to expect.
[00:49:03] Speaker B: The question is, if the officer sees Alex, will they still respond?
[00:49:10] Speaker D: So they do vet human false positives.
[00:49:13] Speaker E: Right.
[00:49:13] Speaker D: So because they can access live monitoring as well, they can make the decision that, hey, these are maintenance guys, or, hey, these aren't people posting in a meeting.
[00:49:22] Speaker A: Yeah, we don't look that suspicious.
[00:49:23] Speaker E: Right?
[00:49:23] Speaker A: That is true. I didn't think about that.
We might have maybe got a fake bike or something and try to maybe take a bike apart.
[00:49:31] Speaker E: Right.
[00:49:32] Speaker A: Maybe like a chop shop activity that might have been more effective.
[00:49:36] Speaker E: Sure.
[00:49:37] Speaker B: So I'm thinking, yeah, that's a good point. If the officer is looking at the video and deciding whether this is a real threat or not. Really?
[00:49:44] Speaker C: Whether that's a known.
[00:49:45] Speaker B: Yeah. He might decide not to respond.
[00:49:47] Speaker A: That's very true.
Which would be actually a good decision. Yes.
I don't think we're just hanging out, so that could be a problem, but we'll see.
[00:49:59] Speaker B: You don't want to.
[00:50:01] Speaker A: Our vehicles are not marked security because we want to show up and surprise them.
[00:50:05] Speaker E: Sure.
[00:50:06] Speaker A: If you see security, they'll run.
[00:50:08] Speaker C: I diagnose, interdict.
[00:50:09] Speaker A: So we don't ever use any security logo.
[00:50:12] Speaker C: The attention to the detail is really.
[00:50:14] Speaker A: Someone would think, a criminal would think, oh, that's just a regular person parking.
[00:50:17] Speaker E: Yeah.
[00:50:18] Speaker B: So is that your vehicle?
[00:50:19] Speaker A: Yeah, that's our vehicle.
What time is it?
[00:50:24] Speaker D: 1046.
[00:50:25] Speaker A: No, how long has it been?
[00:50:27] Speaker B: Took about four minutes.
[00:50:28] Speaker C: Less than four minutes.
[00:50:29] Speaker D: So the alert was triggered at 1041. 1046. So five minutes.
[00:50:34] Speaker E: Not bad.
[00:50:34] Speaker D: Under five minutes.
[00:50:35] Speaker A: Hey, we're just testing the system today.
I appreciate it. Yeah, I know you're busy, man. We don't need to keep you, bro. Yeah, I appreciate it.
[00:50:45] Speaker D: Thanks for your, uh. Under five minutes. That's awesome.
[00:50:48] Speaker E: Yeah.
[00:50:48] Speaker A: Good job. Really great.
[00:50:49] Speaker E: Yeah. Wow.
[00:50:51] Speaker A: This is PK.
[00:50:52] Speaker E: Jason.
[00:50:52] Speaker C: Hi. Cody.
[00:50:53] Speaker A: He's one of our supervisors in the field. This is obviously a very important neighborhood, and a very important person in that neighborhood is what you need. So he's our guy here.
Thanks, bro.
[00:51:05] Speaker E: Have a good day.
[00:51:05] Speaker A: Be safe. Your beard looks majestic.
[00:51:09] Speaker B: That was great, by the way.
[00:51:11] Speaker E: Yeah, great.
[00:51:12] Speaker B: It's great to see this in action.
[00:51:16] Speaker A: It's amazing because, again, the average response time to a critical incident, a shooting is 14 and a half minutes important.
[00:51:23] Speaker E: Right.
[00:51:23] Speaker A: And the average commercial burglary, the average crime is under seven because every criminal knows they have at least seven minutes.
[00:51:30] Speaker B: Seven minutes. And we actually, I've seen that we.
[00:51:32] Speaker A: Beat it by two minutes.
[00:51:34] Speaker E: Right.
[00:51:35] Speaker D: So we interjected, we can get the arrest prosecution, and that's what locks down these properties. Awesome.
[00:51:41] Speaker A: You want the next property?
[00:51:42] Speaker E: Yeah.
[00:51:42] Speaker D: Let's go test a different district. So the way that the security company works is they have different districts all throughout Portland.
[00:51:48] Speaker E: That's right.
[00:51:49] Speaker D: Right now we're in what's called the Pearl district. So we're going to head down into a more saturated area in downtown, and we're going to be testing out another site with a different officer response.
[00:51:59] Speaker E: Awesome. I love it.
[00:52:10] Speaker D: All right, so this is our second property that we'll be testing Echelon's response from. This is the PA living building. This is part of a larger property management group that has deployed our services on multiple of their properties. So today, now that we're in a different district, it should be a different officer. So we're going to go over and we're going to trigger the loitering analytic. When we first took over this property, we were getting calls all day, every day. And one of the things that happened was when one of the employees of the property was interacting with a houseless person and trying to report to dispatch there was a retaliation attack that happened. And so now this completely eliminates that uncomfortable interaction, and it drives the safety of the property up as well for the employees and for clients. So let's go get an idea of what takes in downtown.
[00:53:03] Speaker E: Great.
[00:53:06] Speaker D: And it's going to be right here. Okay, so one of the bigger issues that we've had, and you can see here.
[00:53:11] Speaker A: Right, so it's an exit.
[00:53:12] Speaker D: There's a lot of primarks. This is actually an entrance. It's a bike room. So this is where they store their bikes. So not only is it a great spot to stay dry and to loiter and use drugs, but it's also a lot of high value items in here. And so you can see all the different, all the different primarks that are taking place here.
So we can go ahead and all just stand in here.
[00:53:36] Speaker B: That's the camera?
[00:53:38] Speaker D: Yes, exactly.
So right now we are detecting, we have video analytics that are tracking each one of you.
[00:53:47] Speaker A: Oh, wow.
[00:53:48] Speaker D: And here, in about 15 seconds, you'll be triggering the analytics which just now was triggered. So the officer should be getting a response.
[00:53:59] Speaker A: Now that makes sense. I mean, getting in and out of a door shouldn't take more than five to 7 seconds, even with a bicycle. Really? Maybe ten at the most.
[00:54:08] Speaker B: So your default is 15 here at this location.
[00:54:11] Speaker D: It is for 15 seconds before it triggers the notification. And now, PK, if you come over here, what's really cool is even though you're out of the zone, stay right there. Your analytic is continuing to track you in red.
[00:54:24] Speaker E: Right.
[00:54:25] Speaker D: So even though you've left the zone, our officers can be confident that as you're leaving, that you are the one that still created that incident. So, again, we're able to track real time. So if the officer is monitoring the live camera, they'll be able to see you going from.
[00:54:42] Speaker A: And again, remember, this is a less residential area. Look around. There's no eyes.
See how there's no eyes on us? Because this is all commercial. People are up in their buildings working. So this is a prime area to break into a room and just steal a bike.
[00:54:57] Speaker B: You had a lot of issues here, it looks like.
[00:54:58] Speaker E: Yeah.
[00:54:58] Speaker D: So we had a lot of loitering issues and tents set up. And so what this does is it gets us early detection on people who are trying to camp or trying to set up tents and so we can intervene, potentially offer resources.
[00:55:10] Speaker B: So you got set up at different zones here?
[00:55:12] Speaker D: Yes, exactly.
[00:55:15] Speaker A: We are at Terry pumps. There you go. And that's amazing. That really is amazing.
[00:55:21] Speaker E: Yeah.
[00:55:22] Speaker D: So two minutes, and we have an officer on site.
[00:55:26] Speaker A: How's it going?
[00:55:28] Speaker B: It's this guy. It's this guy.
[00:55:30] Speaker A: He was recording.
[00:55:31] Speaker B: He was. No.
[00:55:32] Speaker E: Do you live here? I don't.
[00:55:34] Speaker A: This is.
[00:55:37] Speaker E: Jason.
[00:55:38] Speaker B: PK. PK, nice to meet you.
[00:55:40] Speaker A: How's it going? Where were you at when you got alerted?
[00:55:43] Speaker F: I'm actually with a gentleman on the other side here. I'm giving him a couple of minutes.
[00:55:46] Speaker E: Okay.
[00:55:47] Speaker A: Take your time. You can go do your thing, bro. So you were already dealing with something?
[00:55:51] Speaker F: Yeah, I was over here on lot 49, dealing with an individual just rolling up a cigarette, hanging out on property.
[00:55:58] Speaker A: Trespassing.
[00:55:59] Speaker E: Okay.
[00:56:00] Speaker C: Trespassing on property.
[00:56:01] Speaker E: Okay, awesome.
[00:56:04] Speaker A: So you got the alert. You were able to still do a little bit of work with him and then break contact, come over here.
[00:56:10] Speaker F: Knowing how close I was, I figured I could do both.
[00:56:13] Speaker A: Okay.
[00:56:14] Speaker E: Wow.
[00:56:15] Speaker A: That's awesome.
[00:56:15] Speaker E: That's amazing. Yeah.
[00:56:16] Speaker D: Great response time.
[00:56:17] Speaker E: I know.
[00:56:17] Speaker D: Less than three minutes in this district.
[00:56:20] Speaker F: Specifically, with the SOS going on, generally, there's an officer within five to five minutes, if not shorter. We can be here.
[00:56:27] Speaker A: Awesome.
[00:56:27] Speaker B: That's great.
[00:56:28] Speaker A: Thanks, Mark.
[00:56:29] Speaker E: Appreciate you, bro.
[00:56:32] Speaker A: Let us know if you need anything.
[00:56:33] Speaker F: No, no worries.
[00:56:34] Speaker A: All right, brother.
[00:56:36] Speaker E: Wow, that was great.
[00:56:36] Speaker A: That was awesome.
[00:56:37] Speaker E: Wow.
[00:56:38] Speaker A: The weird thing is, during that entire time, not one person walked by. Again, these high, dense areas, these commercial areas, business districts, when no one's walking around, you need to have community engagement. What makes an area safe is increasing community engagement. So having that district officer walking around, engaging every single person on all the properties, this is what is going to make people feel like, I need to go somewhere else.
[00:57:03] Speaker E: Right.
[00:57:04] Speaker A: If I'm going to commit crime, I should go somewhere else.
[00:57:06] Speaker C: Technology in action.
[00:57:07] Speaker A: Technology in action.
[00:57:09] Speaker E: Actual action.
[00:57:09] Speaker B: Actual things happening.
[00:57:10] Speaker A: Yeah, this is great.
[00:57:11] Speaker C: Providing the action of the intelligence.
[00:57:13] Speaker B: It's what we talked early in the morning, how things happen, response times, and now we are seeing that in action.
[00:57:18] Speaker A: I'm just surprised. My guard.
[00:57:20] Speaker B: Yes, he showed up.
[00:57:21] Speaker A: He showed up and what, two, three minutes?
[00:57:23] Speaker D: Yeah, three minutes.
[00:57:24] Speaker A: I mean, that's amazing.
[00:57:26] Speaker B: And over there in less than five minutes.
[00:57:28] Speaker E: Yeah.
[00:57:30] Speaker D: The bigger difference, too, between this district and the Pearl district is that the Pearl district is a lot of roving, so they're doing a lot more driving as well as walking. And downtown is a lot more of a walking district. So the fact that he got here in three minutes was incredible. He could have been six blocks away.
So really great response time down here.
[00:57:49] Speaker E: Wow. Awesome. Yeah.
[00:57:51] Speaker C: Technology in action, enabling, solving problems, security guys, focus on what he or she could really impact and contribute for the community.
[00:58:01] Speaker A: Exactly.
[00:58:02] Speaker E: All right.
[00:58:03] Speaker D: So right, next, I think we're going.
[00:58:05] Speaker A: To hit up a smart community.
[00:58:06] Speaker D: I think that's all on the agenda.
[00:58:29] Speaker A: So we're here with, again, PK Gupta, Jason. No, Adam Scheider and myself. We're on the ride along. We just finished up on some properties that are running the Overwatch program, or Overwatch services provided by Meg, distributed to different clients by Sentinel. Overwatch services, Adam Schneider here, the COO. And we're here to meet with Hurley development. They're a development company in the northwest territory here in northwest part of America.
They have a large swath of land they're going to be developing, and they want to make that entire development smart. So all the buildings, both residential, commercial, everything connected and working together. So we're going to meet with this team and see what the possibilities are. So today we've been seeing these other sites, and now we're having that larger discussion of what it looks like to do in larger platform, multiple buildings. Right. Maybe a city, public spaces. Exactly. So that's kind of what we're doing today.
[00:59:29] Speaker E: Hello.
[00:59:30] Speaker C: Hi.
[00:59:31] Speaker A: We have a meeting, I think Nick and Jason.
[00:59:34] Speaker E: Okay.
Hey, good to see you. Yeah, how you doing? Good.
[00:59:39] Speaker A: Yuri. Nick, good to see you, bro. Yeah, why don't I choose you to Echelon and SOS. This is Adam, coo sentinel overwatches. Hi, PK CEO, make pleasure meeting you.
Yeah, sure, that's perfect.
[00:59:55] Speaker G: He's on the phone right now, but if you write in.
[00:59:57] Speaker A: Yeah, perfect, thank you. Hey, Jason.
[01:00:00] Speaker F: Hey guys, what's up?
[01:00:01] Speaker A: Hey Jason, thanks for meeting with us today. I know you're really busy, I appreciate it.
[01:00:06] Speaker F: Absolutely.
For you, anything.
[01:00:09] Speaker E: No.
[01:00:12] Speaker F: So glad you guys come out and chat about this. And we love what you guys are doing on the podcast. And it's really interesting because the stuff you guys are talking about is the same kind of stuff we're dealing with here. We're dealing with it from the developer side. So as a developer, we're looking at a community like this and we're saying, what would it look like for this to be a connected community?
We might call it a mini smart city. Micro smart city is probably what it is. But you do have 100 acres here, 107 acres.
You've got a lot of residential living up in here and here and down here through here you've got industrial, a lot of areas where what you guys run into, I'm sure is concerns about security, about products, about ip, everything from biotech, it could be somebody that's manufacturing small specialized machinery, all different kinds of stuff where they have concerns about security.
[01:01:27] Speaker E: Sure.
[01:01:27] Speaker F: And then through this center core you have all different applications of retail, restaurant, entertainment, hotel, and even some living in here and some podium buildings with inline, you know, retail and restaurant underneath. And so with all these people with over 2000 units of residential, we've got some active senior living partners here.
We have these views of the Columbia river that you can't see off here. And the gorge is right over here. This is kind of the gateway to the gorge. So when we build something like this, we're saying how can we be connected to create lower energy consumption, moving toward net zero as much as we can?
How can we encourage offsets using solar?
There's a lot of wind through here, how can we encourage, we've talked with Yuri about wind power, just making it smart to solve the problems that they have, but also solve the problems that we all have for our kids and our grandkids with climate change and just where we can taking down things like emissions. There's some users that we're talking about that might come in here that are very powerful. They could bring a million guests a year. Just between a few users there so again, what does that do for traffic? Now you've got to put a garage there. Okay, what does that do? How do you manage that? Well, you want to be able to bring that online. The intelligent back end of it is there already. The infrastructure we want to own and partner in. The dark fiber that goes all through here and connects and it's coming in and out. It has more than one direction. If it goes down on one side, it's covered on the other side. Right.
And then all of know, Alex, where you guys especially shine, all of that needs a ubiquitous hidden protection.
[01:03:57] Speaker E: Safety.
[01:03:57] Speaker F: Security where we don't have drones flying over. They don't know it, but it's happening at all times.
[01:04:07] Speaker C: Non invasive, pervasive security and safety.
[01:04:10] Speaker F: Yeah, it just blends in. We want to have safety officers that are kind of walking around, but they're there even just to help, just to say, hey, what you need to find your way.
[01:04:20] Speaker E: Let me help you.
[01:04:23] Speaker A: Community, public safety.
[01:04:24] Speaker F: Yeah. We want to promote community and yet have it be safe. And part of what makes a community feel good is that it's a safe place.
So, yeah, that's kind of an overview.
[01:04:38] Speaker B: Of what's going on.
[01:04:39] Speaker E: Yeah.
[01:04:40] Speaker A: Interesting. I'm interested in getting your thoughts on that.
[01:04:42] Speaker B: Yeah, no, so this one, you described it. I think you're right. It's smaller than a smart city. Like maybe a smart community. That's what you want. And you can have it on fiber, you said, but you obviously have to have wireless coverage. So I'm thinking wifi. You're thinking people nowadays are thinking even a private 5g, providing a 5g network, that becomes more.
[01:05:05] Speaker F: And small cell partners.
[01:05:07] Speaker B: Small cell partners over there.
[01:05:08] Speaker F: Yeah, too.
[01:05:09] Speaker B: So private 5g networks, actually one of the biggest applications on private 5g networks, which people are deploying in communities like this, is video analytics to put on top of that.
[01:05:18] Speaker E: Right.
[01:05:19] Speaker B: That's very easy with cameras there. And you run these services on top of the network. Initially, you can start with Wifi also, this is all very well aligned with what people are trending towards. And then you put different sensors also.
[01:05:32] Speaker E: Right.
[01:05:32] Speaker B: You put sensors for different things.
[01:05:34] Speaker E: Yeah.
[01:05:34] Speaker B: It basically becomes a smart community.
[01:05:38] Speaker C: One of the issues with the sensor is that it always required power source, reliable power source, because sensors, unless it's 24 x seven, it doesn't really mean much to you.
But with the street lights you have always on power. That's where you like to put your fill. Sensors, hotspots, cctvs and network endpoints. That way you'll be able to collect all the data and having that visibility, real time visibility, would give you the very reliable sort of infrastructure so that you'll be able to learn and adapt and make decisions as we move along.
[01:06:16] Speaker A: Both of these gentlemen, PK and Jason, have worked on and currently are working on smart cities. So this is stuff that they've already completed and done. This is anything new for get. They're titans in their field. They're very humble. They don't talk about their background.
[01:06:33] Speaker B: But one project that you might find interesting is there's a couple of smart city projects I'm involved in. One is in Las Vegas. They're big city, obviously, but they have a couple of blocks. They're marked off as a smart block, smart city project. And what they're doing there is. So Cox cable is providing the infrastructure.
[01:06:49] Speaker E: Right.
[01:06:51] Speaker B: But the hotspots are street lamps.
[01:06:54] Speaker E: Right.
[01:06:54] Speaker B: So the street lamps they're deploying in this couple of blocks, few blocks. The lamps, besides, obviously providing light, they have cameras installed in the lamps.
[01:07:03] Speaker E: Right.
[01:07:04] Speaker B: So the lamps provide coverage of the streets. And they got 5g modems in those. So they're also hotspots. They are 5g hotspots. Now, you could have them as Wifi hotspots, but 5g hotspot gives you much better coverage.
[01:07:17] Speaker E: Right.
[01:07:17] Speaker B: And much better bandwidth. So they're doing that. They're much more advanced, they're much more well funded, and much further along. But a small suburb of Atlanta, Peachtree, is also launching this thing right now. So for Peachtree, the local town, we engage with them through a partner. And what they're doing is they just want to monitor a few public spaces.
[01:07:38] Speaker E: Right.
[01:07:39] Speaker B: They want to monitor, like, parking spaces for parking and see how it's been used so that there can be some indication on the app that this slot has so many empty spots, you can just go in. So you don't have to put in sensors like they have in PDX airport.
[01:07:54] Speaker E: Right.
[01:07:54] Speaker B: Every slot has a sensor. Those get expensive, but with camera coverage, you can still get some good information.
And then they also use it for. Yeah, they have a public space for people gathering, and they want to maintain, they want to monitor the crowd level and any anomalous behavior, like crowd, say some fighting, erupting or something.
[01:08:16] Speaker E: Right.
[01:08:16] Speaker B: So you can track behavior of people if they're running and jumping instead of just walking. So you can track those behaviors and then kind of give an early indication that something might be building. So those kind of things are happening right now.
[01:08:28] Speaker E: Right.
[01:08:28] Speaker B: So with your project, you're starting from grounds up. It's a great opportunity to bring all that in and then define that, eventually creating more value for the residents.
[01:08:38] Speaker E: Right.
[01:08:38] Speaker B: That's what you.
[01:08:41] Speaker F: Need to. The users need to see the vision, and then the residents and the guests come because the right users are there, so they need to see the value.
[01:08:52] Speaker G: Nick, I think you should talk about the master plan, like, a little bit. Know, Jason gave the high level of how you guys are innovating.
From an architecture standpoint, how would you see employing this advanced controls and AI solution?
[01:09:13] Speaker H: Well, I always like to kind of step back a little bit, because, just to give context, this is an existing rock quarry, and right now it's kind of a scar on the face of this area here. But this has a huge potential and opportunity, and that's what Hurley sees in it. That's what we see in it. And so for us, as we develop this, how do we fold into the existing fabric, and how do we deal with all the edges around here and make it seamless, not just like, okay, here's a development that just kind of popped out of nowhere. How do we blend in? From a physical standpoint and built environment standpoint, but also from a social fabric standpoint, how do we blend in? How do we engage all the existing communities here to really see our development as a value add versus just another developer coming in, breaking in the dough, and really doing what they want? We want to see this as one. We're raising the value of their home, but also as a way to really add value to their social, from how they shop, the way they think, and they really see development in a different light.
We want to see ourselves not just as any typical developer. We want to add value to your community, not just by building and adding commercial and higher rents that will add value to their home, but what kind of value do we really add to their lives? And we look at it from kind of building community again, it goes back to that whole edge condition from a social standpoint.
[01:10:49] Speaker G: Jay, as Hurley's developments director, development, I was going to say what is the most important pieces? Because I'm sure you've heard a lot of different pitches of a lot of different technologies, and we feel very privileged to have PK and Jason in here.
[01:11:09] Speaker F: Yeah, that's a great point.
It's great to have. We're very interested in controls for HVAC. That's key because we can immediately. It's kind of like the old led days where you went into a warehouse and I just saved you 70% on your energy bill, and it'll pay itself off in a year and a half, and that's kind of the age that we're in now with HVAC.
[01:11:37] Speaker G: But I think beyond just looking at one dimensionally, the cost per kilowatt hour, I think communities are looking for more reliable, decentralized energy, so that if the grid goes down, they know that their community is going to be powered. Especially if you've got assisted living or any older population where heat and cold play a factor. And then also if you're looking at light industrial data center or medical communities here, you need 24 x 724 dual. So I think you should talk about microwind, which not too many people hear about.
[01:12:22] Speaker C: I'll hand it over to HVAC.
It looks like every quarter scientific community is coming up with a sort of more intelligent way of monitoring and controlling.
Well, down in this bay area, we are working with the G company to come up with their next generation of data centers, how you control the next generation data center, so that it is net zero, as close as to it. So they are even coming up with their own rec designs, own coolant designs.
In California, the cost is high. And even in the Mountain View area, you're not allowed to expand any more data centers unless you build your own power plant.
[01:13:12] Speaker A: Really?
[01:13:13] Speaker F: Oh, my gosh.
[01:13:14] Speaker C: So they are trying to come up with a new, almost net zero like data center models. Of course, the computer needs to run, but new coolants, new design, new aerodynamic design of the racks, and the new design of the whole data center room so that minimum energy would be used in cooling it down. So that's one thing. So, Hvac control. Yeah, if you could design it, that's best. But once it is designed, there's always a new way to retrofit or make it better, deliver, better cop.
With that said, I've been working with some of the mechanical engineers of how to make more reasonable and practical wind turbine. I don't like those huge thing that generates a lot of vibrations and noise. If you put in the. In the river about 60ft around, there's nothing alive.
Sea animal doesn't like those vibrations. They stay away from those things.
People doesn't realize that. And in Korea, in the Teju island, in the southern tip of island, they put 25 of them. If you dive into it, it's all white, no living things, not even sea sprouts. So I saw that about six, seven years ago. And I start talking to the mechanical engineers. You're the mechanical guys. Is there any way for us to build a better wind turbine? And when you really think about it, what we have been trying from the mechanical engineering perspective, is that we try to reduce the speed into manageable speed. Take an example for a car engine runs about 7000 rpms, but what you get is two hundred k the speed at 160 miles. Right.
We've always trying to reduce it down to the manageable speed and velocity, because.
[01:15:29] Speaker A: It'S about current in order for storage of electric.
[01:15:32] Speaker C: But let's look at the electric. It should be different when you deal with the nature. The wind is like 6 miles an hour, but I need to generate it. For my turbine to run, it needs 20 meters/second of power. How do you do that? We need to reverse the whole mechanical engineering perspective. So I run into this interesting old retired professor who used to design torpedoes for the military. And his job was try to try to come up with a design of the turbine in such that it would stay, first of all, noiseless. It would stay straight lines. That means maintain the consistent current.
[01:16:18] Speaker H: Yes.
[01:16:19] Speaker C: So he's the guy who has patents all over the world. A six patent regarding torpedo design, turbine design that would give you noiseless speed and consistency without tailspinning. So just go straight line. So I asked him, hey, I've dived into this water. I see just whitening of the whole area because of the wind turbine. Can you think of something else? He said he's been talking to this oldest, the military guys, why don't we use this, turn it around and use it. A wind turbine. I funded him, so he come up with this interesting design. So we got the patents and everything, and we build several prototypes. We only require five meter per second wind to generate power.
[01:17:08] Speaker A: That's amazing.
[01:17:08] Speaker C: It's like, as long as that's more than five meter per wind, if you see the trees moving, we could generate the power. So we installed it on school campuses, we install it on some of the islands, so that we'll be able to provide enough power for the whole resident.
[01:17:28] Speaker A: So that island is net zero?
[01:17:30] Speaker C: Yes.
There are three islands in Korea that is purely net zero.
[01:17:34] Speaker A: I cannot wait to visit you with.
[01:17:38] Speaker C: Turbine about this big. About 20 of them.
[01:17:41] Speaker F: 20 of them. How many miles in circumference or in diameter?
[01:17:45] Speaker C: Oh, that's not a huge island, but that's only 200 residents. Okay. Fishermen living in a small community. So we put 20 of them to provide the power.
[01:17:57] Speaker G: And you don't need, like with big wind, you need these massive cranes with this?
[01:18:01] Speaker C: Yeah, you don't need the massive cranes.
I just carry one of the turbines myself. The magic is in a six rpm. Our design is such. You would turn into 60 rpms with enough torque to run the dynamo.
That's the magic of our dynamo.
[01:18:20] Speaker A: The dynamo, yeah. And do you have the patent for that?
[01:18:23] Speaker C: Yeah.
[01:18:23] Speaker A: That's good.
[01:18:25] Speaker C: So those are the things. But some of the people in the overseas, they find it very difficult to believe because you have to turn everything you've knew about the mechanical engineering upside down.
That's the reason we build it ourselves. We didn't want to build it ourselves, but we just wanted to build a concept model and show it to other world to catch up. But they didn't, so we had to build it ourselves and install it with the Korea power to help some of the islands down in southern part of the Korea. So that's where we are. We haven't been able to scale out because we need to find the right partner to mass produce the turbines and the gearboxes in the epicentric gearboxes and all. And the interesting thing is, with the highway patrol office, we are going to put some of them because the trucks moving by generate enough wind for our turbine to start generating.
[01:19:26] Speaker A: That's amazing.
[01:19:27] Speaker C: Yeah. That's something that we are basically wind tunnel.
We are working with this company, and we are trying to convert it into a voice activated interactive kiosks. You don't need to touch this anymore because senior people have difficulty.
We are working with the senior residents, and they can order their dinners using this. So we're trying to provide a more voice activated version of it so that have a just conversation. With this kiosk, you would validate and generate the recipe or the order for you and connect it with the robot. Delivery robot in such a way that once she sits on a table, delivery robots would recognize her and deliver the dinner. Exactly. Dinners for her or medicines for her at that table.
[01:20:23] Speaker A: I have a son. My oldest son has autism, and he's permanently disabled, so he's low functioning autistic, and the school that he attends lost track of him. He got a mile away down the street.
Someone's supposed to be within eyes and ears, within 36 inches of altime. And one of the contractors and a landscaping company left a gate open.
He escaped during recess. He made it a mile down the road, and an older lady found him in her car.
Had she not found him, he possibly could have died. This was only, like, three months ago.
And so that's what I'm thinking about when I'm talking about a safe place. I'm talking about tracking children. It's making sure that my son. That I know where my son is. Even if I put a tracker on him, some type of bracelet. But there's some type of monitoring system where someone who has. If you can sell that type of safety. There are a lot of people that would pay very well to live in a community where they could track their kids, especially kids with any type of accessibility issues. Even a voice, because my son doesn't speak, but when he does speak, his voice could be activated and they could find him anyway. We could talk about this forever. These guys know more than I will ever know. I should have got a PhD or some tech field. I don't know.
So interesting. I learned so much. Every time I'm with you all. It's a privilege.
And we don't want to take up your entire day. I think that the plan is we might go view the site and just kind of.
[01:22:09] Speaker B: Is that site right now. You're sure?
[01:22:12] Speaker E: Yes.
[01:22:13] Speaker C: Okay.
[01:22:18] Speaker A: Well, thanks for having us. I know we took up a lot of your time. We appreciate you letting us come out.
[01:22:22] Speaker F: Love having you guys out anytime.
[01:22:24] Speaker A: You're always welcome.
[01:22:26] Speaker H: Yeah.
[01:22:27] Speaker A: All right, well, we'll get out of here.
[01:22:29] Speaker C: Thanks for the time. Appreciate it.
[01:22:32] Speaker A: Well, that was a really great meeting. I think that we all kind of learned a lot. Really impressive. The development. I'm thinking that we go there and check this site out.
[01:22:40] Speaker E: What do you all think? Yeah. All right, let's go. Roll.
[01:22:43] Speaker C: Let's go.
[01:23:01] Speaker A: Awesome. So we just arrived. We're going to actually get out and check over this development site from a vantage point and let's see what we find. Yeah, I think Jason and Nick are up here.
[01:23:15] Speaker E: Sure.
[01:23:16] Speaker C: Let's go find out.
[01:23:17] Speaker A: This is supposed to give us a good view of the development site. Glad you wear your hiking shoes, Yuri.
[01:23:28] Speaker F: Let's give you an idea of the scope up here.
[01:23:32] Speaker H: Yeah.
[01:23:32] Speaker F: This area up here is all going to be around that corner there.
[01:23:41] Speaker A: So when it comes to developing something of this site, what's your overall budget? Overall development budget, townhomes, your multifamily residential units, condos.
[01:23:53] Speaker F: Great question.
[01:23:54] Speaker E: Yeah.
[01:23:54] Speaker F: All in. This is going to be between a billion and a billion and a half dollars total.
All in.
That can increase as time goes on.
[01:24:09] Speaker E: Right.
[01:24:09] Speaker F: Because it can take three, five, seven years. By the time you're starting to phase.
[01:24:13] Speaker A: When you're phasing it out, you're landing and delivering buildings.
[01:24:17] Speaker D: I have a question. From a security standpoint. In phasing, at what point do you think you want to utilize either tech and or physical security? Are you required to up here in the state of Washington? Once development starts, what does that look like on the construction side for security services.
[01:24:35] Speaker F: Yeah, great question.
So from our perspective, security is always important. Whether it's required or not, we're going to have a presence. But to what level the presence is there and to what level you can kind of afford more technology and better technology depends on the scale of the site. So as we get this all built out, you start to see, okay, this alone needs to have a more tech heavy presence to be able to. Because let's talk about money. You can't save money by putting a bunch of full time people standing around watching for something to happen.
That's actually much more expensive than having the right tech in place.
That's signaling to people who are out in this area anyway, right? Because they're getting. Guys like your companies are talking to the tech companies right over here, talking to the Vic, the Vancouver Innovation center right over there. You're talking to all these people and going, hey, let's get in on this together. Which is building the infrastructure for you guys, that there's always guys driving around, right. They're right down the road.
If there was a call, they're only six minutes away, whatever it is. So our hope is as soon as we have a bunch of people living here, we're taking care of them with security.
[01:26:10] Speaker E: Yeah, love it.
[01:26:12] Speaker A: Really a great plan. You got your phasing in place. Sounds like you got your financing in place.
[01:26:18] Speaker F: Well, we're always looking for partners.
[01:26:20] Speaker E: Yeah.
[01:26:21] Speaker F: Anybody that wants to get in on.
[01:26:23] Speaker A: This kind of project. Yeah, absolutely.
[01:26:26] Speaker F: We're always interested in talking.
[01:26:28] Speaker A: Yeah, deep pockets and big hearts. That's what you're looking for. That's always the best investor, isn't it? Deep pockets.
Impatience. Yeah, of course. Well, thanks for showing us this site. I really appreciate it. This is great. Yeah. You want to head back?
[01:26:42] Speaker E: Wonderful. Sure. Okay, great deal.
[01:26:44] Speaker B: Thank you.
