Disaster Intelligence: Do THIS During An Emergency

Hello and welcome back to another edition of TAC Intel Tuesday—that's Tactical Intelligence Tuesday. I'm Mike Shelby, former Soldier and intelligence contractor, and I'm headed back to Texas right before this hurricane hits. Hopefully, I get home before it starts raining. I said, "You know what? This is a really opportune time to talk about disaster intelligence," and that is the topic of this week's episode. I'll break down what local intelligence operations could look like during a disaster. This case study is Hurricane Florence back in 2019 in North Carolina, but you could apply this to earthquakes, tornadoes, or whatever kind of natural or man-made disasters you're dealing with because the principles are the same. The concepts, the fundamentals are all the same.

Now, before we get into that, I want to start off today's show with a public service announcement. Remember, remember that no matter what you go through in the future—a natural or man-made disaster—what you do, where you do it, when you do it, and who you do it to is driven by intelligence. Intelligence makes your world spin; intelligence is the fuel for your operational engine. I think about it like this: you're in a vehicle—that is your unit. Maybe that's a neighborhood watch, a community security team, a mutual assistance group, or an infantry unit; I don't care what it is. You pull into the gas station, you swipe your card, you put the pump nozzle into your fuel tank, and you start pushing fuel into your fuel tank. That is information—the information we need to collect. What we want to do with that information is turn it into intelligence because we act on intelligence; ideally, we don't act on information. We should be acting on intelligence. So we have this tank full of information, and that is what gets injected into our engine, but before it reaches the engine, it goes through a fuel filter—that is intelligence analysis. That pump is your information collection: your human intelligence, signals intelligence, imagery, open source, MASINT, whatever kind of intelligence-gathering capacity you have. Hopefully, it's all-source, multi-discipline. We want a little bit of everything in there. That's your intelligence collection, and then it goes through a filter, which is your analysis, because we're trying to weed out the particulates, the dust, the dirt, the grime, and whatever else may be in our fuel tank—whatever is being injected into our fuel tank with all that information. The analyst acts as a fuel filter, ensuring that the highest octane fuel is what's actually being injected into our engine because there is no faster way to kill an operational engine than through bad intelligence or no intelligence at all.

If you are preparing for an uncertain future—a natural or man-made disaster—and you have not done a threat assessment of your area, you are really putting the cart before the horse. If you want my 15-minute threat assessment guide, head over to area intelligence.com. I'm going to send it to you for free; I'm going to throw some extra bonuses in there as well. Do those two exercises; it literally will take you 15 minutes, and it's going to help you identify all the major threats and hazards that you will face during any disaster—whatever it is, as long as it's not lizard people—any realistic, feasible disaster, natural or man-made. You're going to understand the threats and hazards that you'll be dealing with, and that's what you prepare for; those are the things that you should be preparing for. If you haven't done a threat assessment, go over to area intelligence.com, sign up for that 15-minute threat assessment guide, and put in 15 minutes; it will amaze you. I teach this stuff all over the country, and it's amazing how many super preppers are out there, and they've got this bunker, they've got this retreat property, they've got years of food, they've got the big Berkeys, they've got the ham radio, they've got all this stuff, and they sit down and they do that threat assessment, and it's like their light bulb comes on, and they say, "I've never even considered doing this before. I haven't understood my local threat landscape until now. Yes, I understand there are some threats out there. I know that we're dealing with this; I know that we have hurricanes, we've got tornadoes, we've got earthquakes, whatever, but I've never seen it listed out like this before," and it really is—in 15 minutes—it's really revolutionary. The last thing I'll say about that: area intelligence.com—go get that 15-minute threat assessment guide.

All right, now for the rest of today's—or this week's—episode, we are going to be talking about disaster intelligence. So this was back in 2019, I believe September 2019, and Hurricane Florence is barreling towards the southern Atlantic Coast—almost Mid-Atlantic Coast—in the Carolinas, predominantly North Carolina. At the time, we were supporting a group called the Cajun Navy. By the way, we're going to try to strike up a deal with them again and continue to do this. And not just Cajun Navy, but listen, any disaster group out there—search and rescue, disaster response, whoever—if you know them or if you're a part of one, reach out to us because we can do area studies; we'll do it for free. We'll do an area study; we can do real-time intelligence support, which is what I'll be talking about today. We do that free of charge for a couple of reasons: A, because we want to help Americans help Americans, and B, this is really good training opportunities for students and for volunteers who want to jump on board, understand the intelligence process, the intelligence cycle, the collection, the analysis, the output, the actual intelligence production—how the sausage gets made—and we want to be able to help people faster and help these disaster organizations run their operations better and faster as well. Shoot me an email or get in touch with me on social media if you want this kind of support.

So it's 2019; we're supporting the Cajun Navy, which is just a bunch of Cajuns and rednecks from the Texas/Louisiana Gulf Coast area, and they are en route to North Carolina in response to Hurricane Florence. They're going to slip in right after the brunt of the storm passes, and then they're going to—they're going there; they're going to stage, set up, bring in their boats, bring in their monster trucks, and start rescuing people. During this thing, there were—there's literally a single mother with her two children on the roof. They helped evacuate—they evacuated themselves, dozens of residents of a senior citizen home. It wasn't the county sheriff doing this; it wasn't the state of North Carolina doing this; it definitely wasn't the federal government doing this; it wasn't FEMA; it wasn't DHS; it wasn't anybody but the Cajun Navy, and that's why I have a profound respect for those guys and what they do. So the first thing that we did is they—one, as soon as they had a memorandum of understanding with the county sheriff that they were going to—this is in Lumberton, North Carolina—the first thing I did was I pulled up a handful of layers on my Google Earth. So number one, I looked at imagery; I looked at the street maps, and I looked at topography. Those are the three types of maps you need at a minimum in any disaster scenario; you need those three layers of your operational environment. Now, the fourth thing I pulled up was a map of the flood plain—a flood zone map—and that is provided by FEMA. You can go to map.fema.gov, punch in your address, and you can download—and you should download—your FEMA flood zones. But the place where the Cajun Navy was supposed to stage was in a 100-year flood plain, and I called Shawn up—where I was talking to Shawn on the phone; he's the director at the time of the C—the Cajun Navy—and I said, "Yo, man, you are right here in a 100-year flood plain—this place where you're planning on staging—and we're talking about a 500-year flood here, so you should probably not stage here," and I will try to find you a better place. So I looked at the topographical map, and what I found was an area of higher elevation that was outside—well outside—any of the flood plains. So there was a Walmart there—a Walmart parking lot—and there was like a hotel and some restaurants and basically everything that you would want for your crew, and so that's where they ended up staging. I think I did mention this on the last podcast—but on the last episode—I think I did mention this on the last episode—but what happened was the—literally the next day they went to that original staging location where they're supposed to be, and it was underwater. So this is one example of how doing just a very cursory terrain analysis, looking at the operational environment, trying to understand the conditions—or the future conditions—that you're going to be operating in pays off bigly, and that's why we do an area study. We definitely start with a threat assessment, but then we should follow up with that with analysis of the operational environment and look at a bunch of other things that are in an area study. That's a topic for another time. So that's the first thing I did. Now, I had a team of volunteers with me, and we all hopped on Discord, and we had a handful of folks volunteer to do the area study.

Now, as a part of this area study, we looked at the six ESAs of the operational environment, which are physical terrain and weather, human terrain, politics and governance, critical infrastructure, safety, security and defense, and economics and finance. And if you want to learn how to do an area study, go sign up for the Area Intelligence Workshop; it's the online version of the one-full-one-day course that I teach. And so we had this information, and we had a really good idea of where important landmarks were and where the neighborhoods were. You know, the next question that these search and rescue teams had—by the way, before—just let me mention one thing here: when we provide intelligence support, we're dealing with intelligence requirements, and intelligence requirements are almost always issued by the commander, and they typically have to do with the future. For instance, you know, just kind of straight off the bat, schoolhouse thing is, all right, where will this enemy unit attack us? Or when will they attack us? Or whatever. Okay, it's future-oriented, and they want to know this so that they can prioritize their own assets and resources to respond to this threat. Now, for us—the search and rescue teams from Cajun Navy—they're kind of—right off the bat—their first intelligence requirement is, where is the greatest need? Where is the—where are the likely areas of the highest volumes of people who will need to be rescued? And I'll just let you think about how you might solve that problem, and then I'll tell you. And the answer is, we looked at a population density map. So we had two overlays: population density and flood zone. And so all we did was we went through and we identified the areas of the highest population density that were in that flood zone, and we gave them that map, and we said, "Here you go; these are the most likely places where you will be active because these are the most—the areas where most of the people live inside the flood zone." So what that did for them is that helped them further refine their operational requirements. That was intelligence that helped them plan for their search and rescue operations. So they were able to stage boats—basically go ahead and start dispatching boats to these areas almost as soon as they got there because there was obviously going to be people there who needed to be pulled out of their homes, pulled off their roofs, pulled out of their attics.

So one task for you this week is—is think about if you've already done a threat assessment—think about your critical scenarios, your critical threats and hazards that you are likely to face during an emergency—and consider what are your intelligence requirements? What future-oriented questions do you have about that scenario? For instance, it might be, at what point after X event—or how many hours or days after X event—will we begin to see looting? Obviously, going to be a huge problem if you have mass looting—something you might want to plan for and react to—and so that is a future-oriented question; that's an intelligence requirement that your—your intelligence efforts—your intelligence section—whatever you have—maybe that's just you; it probably is—that is a question worth answering. Another question might be, at what point will we begin to see out-of-area criminals? So these are not criminals in your neighborhood already existing in your neighborhood; these are people coming from outside the area to your neighborhood for whatever reason. Another intelligence requirement might be, what is the most likely avenue of approach? Or how are these out-of-area criminals going to get into our neighborhood? And that's a very worthwhile intelligence requirement because once we answer that, we can go ahead and begin making plans: all right, how are we going to deter them? Or how are we going to keep them out of our neighborhood? So an intelligence requirement, just generally, is a question or a statement that leads to a decision point. So, you know, how will out-of-area criminals gain access to our neighborhood? What is the most likely avenue of approach for out-of-area criminals? Or at what point, in hours or days, will out-of-area criminals begin to target my neighborhood or my community? Okay, that's—that's the decision point. Once we can answer that, then that leads to some sort of operation, hopefully.

Now we're going to go around the intelligence cycle here, and the intelligence cycle—depending on who you're looking at—I was brought up on the intelligence cycle having five phases. So number—number one, the intelligence cycle always starts with their seat of mission. The commander says, all right, we—my intent is to go here; we're going to do this, and here's the end state. We're going to target these people; we're going to secure this village; we're going to clear and hold; we're going to do XYZ, whatever. For you, that might be, we are going to bug out of our neighborhood, and we are going to end up in our retreat property, and the end state is we are secure from whatever bad stuff happened at the location we just left; or the end state is we are going to provide for security operations in our neighborhood to make sure our subdivision is safe. Okay, whatever operation, whatever mission statement you have—that's what kicks off the intelligence cycle. Now, it's important for us in—in intelligence—to understand what that mission actually is. So if I'm provided with a mission statement, then I can say, hey, this is how intelligence is going to drive this mission; this is the—this is the—I know exactly what kind of intelligence we need to make decisions to plan for operations. And once I produce this intelligence, and the ops guys are equipped with this intelligence, now they can make decisions about what to do on the operation side. Remember, we're—we're operations and intelligence are two different hats. We're dealing with a—you know, S2 or the G2, S3—the two and the three. Okay, regardless of if we're dealing on the company, battalion, brigade, division—whatever echelons above corps—whatever—the two and the three; the two shop is the intelligence shop; three shop is operations, and they kind of hold hands. Part of the military intelligence creed is to find, know, and never lose the enemy. Okay, that is the mission of the—in the intelligence shop—one of the many missions of the intelligence shop. And so if—when operations comes to us and say, hey, we want to know what to do, where to go, when to do it, and who to do it to—only the Intel guys can provide that, which is why I continually say, hey, intelligence fuels the operational engine. If you are not producing intelligence, or if you're not producing good intelligence, then it's kind of hard to drive your operations anywhere. Remember, intelligence is timely, relevant, accurate, specific, and predictive or actionable. Timely being, we—we've got to know it when we need it. The best, most perfect intelligence tomorrow is worthless to me today. Relevant is, it's relevant to the mission. Accurate and specific. Okay, obviously, we've got to be producing accurate intelligence, and we need to be as specific as possible with the intelligence. So, you know, don't just tell me, hey, there's a bunch of Taliban down here. How about, hey, there's like 50 to 75 Taliban armed with AK-47s and RPGs; they have likely pre-mined; they've probably already embedded IEDs on these avenues of approach—i.e., how gun trucks are going to get in there, how patrols are going to get in there. Okay, those—those roads and those—or those footpaths or those passes through the mountains have already been mined, etc., etc. Be as specific as possible with the—with the intelligence that you provide. And then finally, predictive or actionable. And predictive is, we—is we can exploit patterns. You know, like for this—just for a basic example—when I was down in Kandahar, my last tour, there was basically a logistical pattern that emerged: when—when some of the local guys got a delivery of IEDs, it was always like two to three days later, all the roads would be mined. And so there was like a two-to—two or three-day gap. So whenever IEDs arrived, we always knew that there's going to be a huge uptick in IED strikes just like a matter of days later. So that might be something that's predictive: there's a shipment of—you know, 100 IEDs came in; we don't know where they are, but we're obviously looking for them; they've been cached somewhere; and now these Taliban IED sales are going to be going to these caches, pulling out these pressure plates, doing whatever—like pulling them out, getting access to them, or distributing them—and then they're going to start popping up on these footpaths and in these roads in—in the next two to three days. And that's just a basic example of predictive intelligence. And then actionable intelligence: actionable is something that we can act on now or act on in the near-term future. So, for instance, if—once we find that IED cache, we say, hey, we think this IED shipment was taken to this compound because this is the compound of an IED builder, or this is a compound of an IED cell commander, or you know—this is—I don't know—whatever—we just have reason to believe that this IED cache is here; this pressure plate cache is here—and okay, so that's actionable: like we can—we can go there and drop a bomb or do whatever; we can go recover these pressure plates, make sure that they are not out in our battlespace killing our soldiers. Timely, relevant, accurate, specific, and predictive or actionable.

So the next thing after we did—after we identified the most—the areas of highest need—or the likely areas of highest need—we had access to a couple of different tools, and one of those helped to scrape social media. And so what happened was there would be these families or these kids or whoever, and they had Facebook accounts—it was mostly Facebook—and they would be taking pictures and putting on Facebook or saying like, hey, we live at 123 Bakaka Daka Street, and we need help because we're on our roof. Well, we could find those posts by scraping Facebook, and there's a thing called geofencing, so we could look at all the Facebook posts in a certain area and rip those off of Facebook and then send those over—either get in contact with them or—or really what we would do is send them to dispatch, and the Cajun Navy had their own dispatch system, and dispatch will get in touch with those people and make sure that they still need to be rescued, and then the search and rescue teams would go out by vehicle, by boat—however they're going to get there—and rescue those people, bring them off the roofs or whatever, and take them back to safety. Another thing we did was monitor Twitter because there were—believe it or not—there were people talking about conditions, and we did keyword searches; we did geofencing; we did—I think maybe—maybe even—there were some tags—like some hashtags that people were using. And Twitter was not a—a great source of information, but there was this one instance where one of the city councilmen said that the—the dam was about to—was about to breach, and that made its way through social media, and everyone started freaking out because they just had this picture of this giant tsunami of water crashing down through Lumberton and wiping out homes and businesses and bridges and everything in its path. Obviously, this is a huge deal. My thought was, we've got to get these search and rescue teams out of the water—like pretty much now—but I remembered back to the area study, and I said, you know, I don't remember—there's not a dam in Lumberton, so I don't know exactly what dam they're talking about. So I jumped on the USACE website—U.S. Army Corps of Engineers—and they have a—a dam—a national dam inventory. I started looking around Lumberton for what—what dam this guy was talking about, and there was no dam. I started calling around—called around to a couple of people—there was like a gas station right near a bridge, and I started—you know, I was asking like, hey, if you—is—how high is the water? Like, is it up over the bridge yet? Okay, human intelligence here, and they were like, no, there's—you know, it's not wiped out the bridge yet. Well, it turns out that it wasn't actually a dam breach; it was a levee breach, and—and it even turned out that the levee was not at risk of being destroyed, or the water was not going to destroy the bre—the levee; there was like a little trickle of water coming over the levee, and that's all it was. So that was the—that was the dam breach: just water was trickling over the top of this levee, which in itself is not insignificant, but the point I want to make here is the reason why we have analysis is to prevent operational snaps like that. Okay, this is a very good example of how intelligence—or bad intelligence—or misunderstood information can produce operational friction, which can slow down operations because if I had just taken the city councilman's word for it and I called Shawn and I'm like, hey, Shawn, this dam is about to breach; you've got to get all your search and rescue teams out of the water till we can figure out what's going on because there's about to be a huge wall of water come rushing down the river and it's going to flood into everywhere—which is one—one possible course of action I could have taken—instead, I relied on the available intelligence that I had—well, almost immediately—in like one—one to two minutes—I figured out there was no dam at all, and it was only a couple of minutes later that I figured out there was talking about a levee, and then it was only a couple of minutes later that I got in touch with the search and rescue team that was in the area, and I asked them—I basically tasked them—hey, can you go get eyes on this levee and tell me if this thing's about to break? Can you do that safely? They said yes, so they went and did it, and they said, yeah, there's just like a little bit of water flowing over the top, and there's a bunch of people out here looking at it, and you know—it—it's not about to break; it's not about to—the water is not about to burst through. Okay, and then we figured it out: hey, this is actually not this tsunami scenario that was going around on social media. So intelligence can reduce operational friction, or it can multiply operational friction to the point where operations stop—which is almost—well, it didn't almost happen, but it's what could have happened had I not looked at the situation with a finer lens or a clearer lens than what was being provided.

Now, it was also probably a day and a half, two days later, maybe, and all the information about all the rescue requests—all the people posting, saying, hey, I need to be picked up—still just stopped. And the question for a few minutes was, well, have these search and rescue teams saved everybody? And the answer was no; there are still plenty of people who needed to be rescued, but their batteries on their cell phones were going out because the internet was down, the power was off, and all they had is their cell phones and a cell—cell signal, and—and then that was it; there was no more connectivity, which is a really good reminder that having all these whizbang tools and having the internet, having instant communications via cell phone is great, but we cannot rely on cellular communications during a disaster, especially where there is no power, because you know all these sources of information are going to run out of juice; they're going to run out of batteries in the phone, and that's it—kiss your stream of information goodbye because it will not exist. Which leads me to my next point: someone asked, Mike, what do you think about the cell phone-based—the cell phone app police scanners? I generally dislike them for a few reasons; I'll tell you why. Let me just say this: if that's all you got, then that's better than nothing, but there are several reasons why I want an actual police scanner and not a—the 5—you know, the 5—whatever cell phone app police scanner there is. Now, the way these police scanner apps work is there is someone out there who has an actual scanner running locally, picking up a local radio signal, and they're piping the audio from their police scanner into their computer, and then they're uploading the audio to the app server, which is rebroadcasting that information out to users. And again, if that's all you got, that's better than nothing, but there's a few problems with that. Number one: if the—the primary source police scanner goes down for any reason—lack of power, lack of internet, whatever—then the whole thing goes down. So it won't matter how great the app is; if the primary source is down, you ain't got nothing. The second reason: I have seen where those rebroadcasts are delayed by 15 minutes in 2016, and '17, and '18 when those police scanner apps got really popular because people were listening in to the Black Lives Matter—I guess going back to 2014 and '15—but they were listening to the Black Lives Matter riots and the—all the Trump protests and counter-demonstrations and all that stuff. Police requested a delay be put on there, and so, of course, like Broadcastify and other—these other scanner apps complied—for the most part. In an emergency, I do not want to rely on 15-minute-old data during a riot or during a disaster. 15 minutes might as well be 15 years in some instances. So I want as fresh as possible, which means I need a primary source; I need a direct source of information, which is not my scanner app. Now, the third reason—I kind of already mentioned this—is this may not be as big a problem on today's phones, but back when we were doing this—really getting into the heat of it in like 2016, '17, '18—and actually back when we did the—the Ferguson riots—was that 2014 and '15—you could not run the scanner on your phone and also make a phone call or have any other apps open. So the couple of guys that said, oh, I'm just going to bring up Broadcastify or Scanner 5 on my phone and then type in what I hear—what—it couldn't work because you—you couldn't run two apps at the same time. Now, maybe that's changed with these newer cell phones, but even if that hasn't changed—which I think it has—but even if it hasn't, you still run the problem of battery life. Your battery is going to run out much faster on your cell phone than on your police scanner. There may be some other reasons, but those are three really good reasons why you—you ought to have a police scanner, especially in an emergency, so you can get firsthand, direct access to the reported information and not get it through social media or not depend on it from someone else's police scanner who—you know—who knows—who knows what's going to happen to that feed during an emergency.

Now, going back to the Ferguson riots in 2014 and '15, we were able to produce real-time information because we had an all-source approach to collection—a multi-disciplinary approach. We gathered as many sources as we could. Now, this is—we call this Operation Urban Charger. If you have a copy of my 2015 book, *SHTF Intelligence*, I talk about this in detail about how we did this, and also in the back of the book, there's an appendix with all message traffic in there. And some amazing stuff happened over the radio that—that we heard over the radio, including gunshots. The media the next day said, oh, we debunked this; there were no gunshots; no one was shooting at cops or whatever during these Ferguson riots. BS, man, because you could hear gunshots coming over the police scanner. We heard it in real time; for sure, that happened 100%. Also, one of the police officers—there was an AR-15 stolen out of the back of one of the police cruisers, and we know this because the cop got on there and said, hey, my AR-15 has been stolen out of the back of my cop car. So I don't know—primary sources, guys—you get the best stuff from the primary sources. All right, so let's talk about the four primary intelligence disciplines that you need access to. You could add others, but these are the four that you really need: number one is human intelligence; number two is signals intelligence; open source intelligence; and imagery intelligence. And I talk about those in detail in some previous episodes; I won't go super into them now, but we'll just start with human intelligence. So we did have some law enforcement officers on the ground in St. Louis and—and in Ferguson—or in the area; they were pretty busy; they were not a huge help, but we did get a handful of text messages from them that were helpful. We also had ham radio operators who provided some secondhand information that they were getting from other people. So we did have people monitoring local radio transmissions; obviously, the police scanner is just a very basic example of like low-level voice intercept type stuff. We had people monitoring local VHF, UHF, and HF, which was not incredibly helpful. Next, we had open source: so we had stuff coming in from social media; we're tracking reporters and some of the riot trackers—you know, like the people—they—the independent journalists who were on the ground doing the videos—and that was back during the days of Periscope. So we're watching video live feeds; we're looking at pictures—that's more on the imagery side—but that was being reported via open source; we're looking at firsthand accounts of people who are there who are live-tweeting or live-posting what was happening there. And then we have imagery intelligence: so we're pulling off the photographs, pulling off the videos, and looking at the landmarks—street signs, business names—whatever was in the background so that we could apply a geotag—basically plot that photo on the map—which brings me to my next point. Listening to this was kind of a perfect storm for signals intelligence—just listening in to what was available on the radio—because at that point in time, there was no—the—the only way that there was radio interoperability between local and county and state law enforcement, National Guard, and federal law enforcement was they are all operating—or at least a portion of the transmissions were in the clear on the same frequency—and so that was just a wealth of information, and that really was the only reason why we could battle track in near real time and produce in near real-time security updates on the overall security picture in the whole area because we had access to all information. So Warfighter 33 was one of the call signs that came over the radio repeatedly, and just through basic context—context clues—we said, oh, Warfighter 33—that is obviously the National Guard commander. Warfighter 33 started telling all those National Guard units to move out to their static checkpoint locations, and this was like 60 minutes—right—we were waiting on—on this verdict whether the jury was going to vote to indict this police officer—we had a 60-minute heads-up before that happened because we were listening to the radio; obviously, they were anticipating rioting. And so 60 minutes out, they started to push out their National Guard units to get into position, and we—we had a 60-minute warning, and then obviously that—that no true bill came out, and then all hell started breaking loose. Now, going back to the imagery—uh—part—reporters were posting photos on social media and on Twitter. At the time, there was a local reporter who posted a picture of the command tent—well, at the time, we didn't—we—we didn't immediately know was the command tent, but there was a military tent—was like a GP medium or something—set up, and it had antennas and it had a satellite dish. And so just a couple of minutes later, we were like, oh, yeah, well, this is obviously the National Guard command post, and we looked in the background of that picture, and there was some red—there was like a retail sign in the background; it had red letters, and it was a Target, and this National Guard command post was set up in the Target parking lot. So boom—we jumped on Google Earth; we dropped a pen; we said, here's a National Guard command post; it is in the Target parking lot; we know where this is. And we did that basically throughout the night. Whenever there was a picture that we could figure out where it was, we just dropped the pen through the photo on there. We had really good situational awareness on what was happening at any given point in the area till like 2 a.m., and at 2 a.m., we were pretty much like, yeah, you know—now we're just doing this for the extra practice, but we figured this out—it can be done. You can take a group of veterans; you can take a group of volunteers—so I think it was like six or eight of us—myself, another intelligence analyst who was stationed in South Korea at the time, and then a handful of volunteers—you—you actually can do this with a small team of people; you can battle track riots, and then by extension, hurricanes and other—you know—whatever. We went on to battle track the Republican National Convention in 2016; we've done it multiple times since then.

So this has been a kind of a crash course, off-the-cuff description of disaster intelligence. This is what we did: just get your human sources lined up; get your imagery and open source lined up; automate as much of—of your open source as you can automate—by keyword, by hashtag, by geofencing—whatever you can do—whatever means you have available—automate as much of that because then that becomes a—it's no longer an active task; it's—it's just passive monitoring, and then you can get active if you're looking for a certain piece of information. You can be active; otherwise, you know, I'm not sitting there banging on my keyboard all night trying to find all these sources; I just—I find the sources; we put them in an aggregation tool, and then we just monitor that message traffic; we monitor that feed throughout the night. If there's anything of intelligence value, we add it to our—our existing intelligence knowledge; we plot it on a map; we—we push it out; we do whatever we're going to do with it. Just use whatever means you have available. This is so important. So listen, there—my last video—or this video before last—I talked about the three mission-essential tasks that you have right now: number one is do your area study; number two, develop