Rod Ammon: Welcome to the CFITrainer.Net podcast. Today, we're talking about the Internet of Things. You're going to learn what that is and why it's an important investigative tool you might not be using. As the technology that saturates our lives and business evolves, data from the Internet of Things is becoming an essential element of a thorough fire investigation. Guiding us in this discussion is ATF supervisory special agent, Certified Fire Investigator Dr. Dawn Dodsworth. She is also an IAAI-CFI. Dr. Dodsworth began her career in a local fire department as both a structural and a wildland firefighter. In college, she interned with the ATF and realized that was the career path she wanted. Dr. Dodsworth served in the ATF's Boston, Louisville, and Seattle Field Divisions. She was a member of the ATF National Response Team and International Response Team prior to her promotion to group supervisor for all arson and explosives assets in the Seattle Field Division and is the ATF Seattle Division Response Team commander. She also is a member of the NFPA 1030 Technical Committee.

For her doctoral dissertation in forensic sciences at Oklahoma State University, Dr. Dodsworth examined the Internet of Things and its use in fire investigations as a source of objective data that can inform fire origin and cause determination. She routinely uses it in her work, and she's here today to tell you how you can put the Internet of Things to work for you in your fire investigations. Dr. Dodsworth, welcome to the podcast.

Dawn Dodsworth: Oh, thank you so much, Rod. I appreciate it.

Rod Ammon: Very grateful for your time and for the research and your dissertation, which I've read some of and really loved what you did. So Dr. Dodsworth, let's start off by defining the Internet of Things. I'm guessing a lot of people haven't heard the phrase or might miss the opportunities it or they present in an investigation.

Dawn Dodsworth: Absolutely. So, just as a quick caveat beforehand is, when I was looking for subjects for my doctoral dissertation, I didn't want to do something that I was familiar with. I wanted to kind of challenge myself and force myself out of the box and to learn something new. So, the Internet of Things basically involves a network of things that are linked together through sensors, communication equipment. It links things among themselves, so sensors to other sensors or devices to other devices, or it could also occur between people and things. So, essentially, it's interrelated devices, machines, and objects with unique identifiers that transfer data over networks without requiring any human interaction.

So, for the layman's terms, we're all very familiar with burglar alarms and fire alarm systems and sprinkler systems, but for laymen and women, we're also looking at things such as Wi-Fi routers, any type of satellite linking, your Alexa app, and your Alexa in your home, Nest thermostats. We have all these smart homes now. So really, when you think about it, the Internet of Things is so pervasive in our daily environment right now that a lot of times, we don't even take into consideration that these devices are present and that, actually, certain other aspects of our lives may be connecting with them, such as our iPhone, or if you have a Fitbit or an iWatch or whatnot. So, it's amazing the amount of data that is out there to tap into.

Rod Ammon: So, just so I correct myself, I always think of it as anything from one thing that might be collecting data to building systems that could be almost running an entire facility.

Dawn Dodsworth: Yeah, these systems are actually included in that. So in your system, you may have a control panel, and then you may have various sensors and devices located throughout the building that are activated or alarm as a mechanism of some type of interaction between the environment and the system or other devices and the system. And there could also potentially be human interaction in the system.

Rod Ammon: Great. So, thank you. I appreciate the clarity. So how does the Internet of Things work? Can you give some examples of how devices are sensing, transmitting, and recording data that can be helpful to fire investigators?

Dawn Dodsworth: Oh, sure, absolutely. So, when we're looking at doing an origin and cause or we're looking at doing a fire scene examination, obviously, one of the goals is we want to be able to determine the area of origin. And the necessity to properly determine the area of origin lies in the fact that our fire cause is going to be within our area of origin. So ultimately, if our area of origin is not properly and correctly identified, then ultimately, the cause of the fire will also be incorrect.

So, when we're looking at these systems and we're considering, there is a number of things that the investigator really needs to consider when they arrive onto the scene. These systems are going to basically... They can be autonomous or human-initiated. It could be anything such as a door or window opening, user commands, motion-activated, smoke alarm activation, sprinkler activation. So we have all these different potential sensors and devices, such as a CO detector, the temperature alarm. We have these environmental control systems. We have these Nest thermostats in homes that are measuring and looking and recording the temperature within your home, and humidity. And we have different water flow alarms and whatnot.

So, when something goes on and interacts with these devices and sensors, it ends up getting transmitted either through Wi-Fi or satellite or cellular network or Bluetooth. There's a number of different gateways that this information is sent, then, to the cloud or a server or some type of remote monitoring system or network, either within that building or ancillary and outside of the building. And basically, then what happens is, there's different output devices, such as sirens and horns and strobes and bells that may be initiated. So, if you're in a large public building and we have a fire alarm that activates, what do you always hear? You always hear the horn, and you see this strobe. But these systems also can be interacted with some type of a user interface such as different apps on your phone or your computer and emails, text messages, and websites. So, the data is constantly being sent out and then sent back between these sensors and devices and the system as a whole.

Rod Ammon: Yeah, it makes for quite a web, shall we say.

Dawn Dodsworth: Yeah, it's actually pretty amazing. I mean, the prevalence of this stuff that is out there is just things that I think a lot of people don't necessarily consider when they're looking at these systems. You want to look and see that, okay, there's sprinkler systems in here. I see a fire alarm, but you're not necessarily considering other devices, such as maybe smart appliances or Wi-Fi systems or whatnot.

Rod Ammon: Yeah, and in the worst case scenario, I would think that for an arsonist, it would be hard to keep track of all of this and all the places where data might be.

Dawn Dodsworth: Absolutely.

Rod Ammon: So, you touched on this when we first started our discussion, but I think your motivation is very important. Can you talk more about how you came to the idea of doing this for your topic, for your dissertation?

Dawn Dodsworth: Sure. As part of the doctoral dissertation process, you want to have some kind of research that's done that has not necessarily been comprehensively done before. So what I did was I kind of started looking at some cases and fire scenes that I had been involved with and kind of looking at different aspects of them and trying to look for something that I thought would be an advance to our personnel out there that are doing investigations, be it something small or something large. And the more I looked at it and spoke with people, I really understood that a lot of people are really familiar with a lot of these systems, but aside from the obvious, looking at sprinkler activations and things like that, they might not necessarily be using the systems to their full potential for the amount of data that can be recovered.

Rod Ammon: Yeah, it's... You had another motivation and a path forward. And I love the quote, I think it's paucity of research.

Dawn Dodsworth: Yeah. Yes. Fire investigation, obviously, we're a science, and we're always evolving, and we're always learning new things, and as we progress, sometimes old styles of doing things are debunked, or information that we might have relied on in the past as being reliable, we've now determined is not reliable. So, what I really kind of wanted to try and do was to kind of bridge the gap that was identified in the National Academy of Sciences report, called Strengthening Forensic Science in the United States, A Path Forward, which was published in 2009.

And basically in there, they were looking at different types of sciences and evaluating the different types of sciences and making recommendations as to how they could be better and where they may have some shortcomings. And really, one of the big conclusions of that report was that fire investigation and the interpretation of fire patterns really needed to be put on a more scientific footing. So, while the NAS report is basically focusing on improving and increasing the effectiveness of various scientific disciplines via their recommendations, I think it's our responsibility as investigators to try to push our careers and our jobs forward by trying to identify the issues where it's been showed that we may potentially fall short or we might be able to increase our efficiency, and try to identify those and try to bridge those gaps.

Rod Ammon: From what I understood from what you wrote, that subjective interpretations, whether you're an experienced investigator or not, this is too important, and there's been too much criticism. Can you speak to that?

Dawn Dodsworth: When we're looking at the issue of potential subjectivity, we're really looking at the understanding and interpreting of something via a individual person's own minds. So, you're drawing from your knowledge, training, and experience, but at the same time, subjectivity can occasionally be problematic, because it's incapable of external validation. A lot of times, especially years past, a lot of answers were essentially the, "It is because I say it is," approach. And also, we have these previously accepted rules of thumbs that were later identified as being inaccurate. So we can run into a lot of different issues there with people kind of coming to their own conclusions of things. And essentially, the more ambiguous the data that's presented to us is, the more likely that a conclusion can be influenced by our own biases that each of us has within us. It can distort and discolor the nature of things by mingling our own fallible nature with it, so we could then be susceptible to such things as confirmation bias and tunnel vision. So really, what I was trying to look at is some way, some type of methodology that can be used that can further create a separation between the interpreter, which is the fire investigator, and the work at hand, which is the fire scene processing.

Rod Ammon: Which is a nice bridge to my next question for you, and that was asking you to summarize what you did in the dissertation and what your findings were. I loved it when you said something about Internet of Things gives us an opportunity with hard data points.

Dawn Dodsworth: Yes. Essentially what I did was I wanted to take an approach in my research, which was kind of a mixed-method study. So, what I was doing was essentially, I was creating a population and a sample pool, and then I was looking at cases that these individuals were directly related to that they had worked on and that they had written reports on. And then I also wanted to speak with them individually, just to kind of essentially get an idea of what their personal experience was, and that's more of the qualitative aspect of the research.

So essentially, what I did was I had sent out a blanket email to the entire cadre of our ATF Certified Fire Investigators and our National Response Team members and our engineers at the Fire Research Lab in Maryland, and I basically asked for them to self-report on if anybody had been involved in any of these larger fire scenes, because that's what my dissertation focused on was the larger scenes, as opposed to residences and vehicles. And so, out of the approximate 120 people that I emailed, I received about 20 responses back from people explaining to me that they had been involved in a case that this information was used. So then what I looked at also was, okay, what was the case, and is the case adjudicated yet, or is it still an open investigation? Because obviously, any type of open investigation, I would not be able to use in my research.

Rod Ammon: Good.

Dawn Dodsworth: So essentially what I did was, I then took my sample, and I stratified it into three sections. I took two ATF Certified Fire Investigators, two ATF electrical engineers, and two ATF fire protection engineers, and they were the people that I ended up actually doing the interview with. And then that interview was transcribed, and I sent the transcription to the people that I interviewed, and I just asked them to check it for completeness and accuracy. And then essentially, what I did was I went through and, in research, we call it coding, so going through and looking at different concepts and different terminology and essentially splitting apart what was being said and looking at the verbiage and relinking it together, forming bonds and connections between the different words. So after I did that, then I kind of mapped that out, and basically was able to pull out some of the color commentary from our experts.

Rod Ammon: Nice. So can you share with us some of the things that you heard from the SMEs you interviewed about the potential of this Internet of Things?

Dawn Dodsworth: Yes, absolutely. Really, when speaking with the six subject matter experts, the different things that came up time and again really kind of reflected things that the investigators at the scene really kind of need to look at and consider. So the first one is the system, the type of system, and the sensor and device type. So, when you are at a fire scene and you're looking at possibly trying to utilize this information, what exact types of systems are present? Is there a fire alarm system? Is there a burglar alarm system? Is there a video system? Which, video is really helpful for us, but as we're going to see in one of the case studies that I'll talk about, the failure of those different cameras can also help you to track back to the area of origin of a fire.

Rod Ammon: Mm-hmm.

Dawn Dodsworth: So, you want to be able to identify those systems. It's going to be helpful to interview the owner of the building, or maybe the maintenance supervisor, or if you're at a large residence, maybe the resident, the owners of the structure, the owners of the house, and just figure out, "Okay, what specific types of systems are present?" And then, once you know what types of systems are present, then you're going to start to think about, "Okay, what are the different ways that this particular system would be activated so that it provides some type of context to the data points that you're looking at?" Another consideration is the construction of the individual devices and sensors, considering how these different devices and sensors can fail within a structure. For example, is the sensor device made of plastic, or is it made of metal? Because obviously, heat is going to impact those and affect those differently. You'll also want to consider how is the system installed? Is it a wireless system, or is the system hardwired into the structure? And obviously, if it's hardwired, then you're going to want to have some kind of an understanding as to the orientation of the feeds that it's wired between the different devices and sensors and between the panel. And you want to kind of understand that, because you might have an activation or some type of trouble signal from a particular device or sensor. If it's a wireless device or sensor, that's going to be a little bit different data point than if it's a device or sensor that is hardwired, because the investigator could just be getting some type of a data point from heat and flame impinging upon the conductors or the cables that are connecting it. So we also want to look at that.

And we want to look at and consider exactly where are these feeds being run? Are they above the ceiling, or are they below the ceiling? Because obviously, once again, that's going to affect the data point that we're getting, and maybe the timeframe within a device or sensor activating versus a wired system having an issue due to thermal effect and then sending a signal. We also want to have some type of schematic as to where these different devices and sensors may be located throughout the building. So what I always do when I go to a scene, a lot of times, I'll have an agent that is basically just trying to track down these data points for me. So, they may go to the owner of the building and ask for a set of blueprints. They may contact the company that may have installed the system and ask if they still have a schematic of the installation. We've had it to the point where we might not have actually the hard documentation, but at least the owner or employee could actually draw on a map circles, kind of like the approximate location, because that gives everything perspective and context.

Rod Ammon: Yeah.

Dawn Dodsworth: Then, we want to make sure that we synchronize all those times, because with each Internet of Thing device and sensor system, there's going to be some type of timestamp with that. So we want to make sure that all the times are normalized, so that we're not comparing apples to oranges. We're comparing apples to apples. So, once again, within that, context is absolutely the key. And then finally, we want to be able to properly interpret and apply the data to our fire scene and try to use that to kind of tell us the story of what went on. Again, it's a data point, and it's different systems and a different perspective for the investigator to take when they're working through the scene trying to reach their conclusion. And it's a way also for testing different hypothesis as to area of origin.

Rod Ammon: Well, you know your stuff so well, you went right through to what I was going to bring up, so I'm wondering if there are other conclusions from your dissertation that you felt were important.

Dawn Dodsworth: So, I think that the conclusions, as far as my research, I think there was a lot of strengths that I think are really positive, and that of unmatched reliability. It's really kind of hard to argue with an actual timestamp. So it kind of gives you this unbiased timestamp or empirical data point to look at. And in that, same thing, the investigator, by using this data, are also utilizing hard times, which are actual times as opposed to a soft time or an estimate of something happening. So you know may have a witness that's giving you an interview, and they could say, "Well, I usually wake up at 5:30 in the morning, and so I was up for maybe about 10 minutes, so I'm thinking that this incident may have happened at 5:40 or 5:45." Well, that's an estimate, but when we actually have a hard time. It's empirical. And, once again, context is important, so all this data points has to be normalized, but it's really hard to argue about that timestamp.

I think that using this type of data from these systems, it assists in the removal of the potential subjectivity, and it ensures more accuracy, quality, and integrity of the data points, and it makes them more empirical for the investigator. The data that we obtain from these systems and these devices can not only assist with the area of origin of a fire, but it also can assist with tracking the fire progression through a structure and even potentially with different types of cause hypothesis. So ultimately, the strength, utilizing this data from these systems can provide investigators with a degree of confidence in their conclusions as to the area of origin.

Rod Ammon: It's a lot going on. Yeah, I mean, there's always a lot going on in a fire investigation. One thing I hadn't thought of until I was listening to you was how much complexity this adds to an already burdensome job.

Dawn Dodsworth: Yeah.

Rod Ammon: It's like, like you guys needed one more thing to try to dig up.

Dawn Dodsworth: I know. It's pretty amazing, but the thing of it is, I mean, I'll give you a perfect example of, actually, utilizing these data points in practice from a remote location. And it just was really fortuitous that this situation presented itself in the middle of my written comprehensives for the PhD program. So I was in my office in Seattle, Washington, and I received a call for a large warehouse fire in Guam. So, literally, it was like a call from the future, because it's across the international dateline. And so, it was a storage warehouse where they stored documents and things for government transfers and things like that. And it was a 33,000-square foot warehouse that had burned.

So, in speaking with the local agents and speaking with the local supervisor, I was basically trying to analyze what type of a response we were going to need to have out to Guam. And obviously, just even thinking about how long the trip is out there, I mean, it's going to be arduous for anybody, but I wanted to make sure that I was sending the proper personnel and the proper amount of personnel for what the scene actually required. So, what I would do was each morning I would get up, and I would draft a long email to the supervisor out in Guam, and I would give him a punch list of things to look at and information to gather for me, and then at the end of the day, it would be kind of like the middle of the day for them, I would get the answers to my questions, or I would get the photographs or the information I needed, and then I would analyze that, and then I would come up with another list.

So essentially, we were able to get really good schematics from inside of the structure, and there was a camera system within the structure that was functioning, and they also had heat detectors in there. And essentially, what I was able to do by tracking the activations and looking at the different camera systems is essentially narrow down the area of interest for us from 33,000 square feet to 1,000 square feet.

Rod Ammon: Wow.

Dawn Dodsworth: And that was literally from like the other side of the world. So, as a result of that, we didn't need to send the National Response Team activation out there, which would've involved 20 or 25 agents and experts going out there. And it would've been crazy expensive for the agency to have to foot that bill for the travel out there. We were able to essentially tier our response to... I sent... Two of my Certified Fire Investigators from Portland went. I had one Certified Fire Investigator from our fire research lab in Maryland go. And I had two engineers from our fire research lab. It was a electrical engineer and a fire protection engineer. And then, we had a number of regular general field agents that are not fire specialists, they're they're just plain old special agents. They actually responded from Honolulu. So essentially, we sent about eight people to assist the Guam field office, and by the time those experts had arrived, we were able to, still working remotely, pull a lot of data points for them, and I was writing up a briefing paper every day and sending it to them so that while they were traveling, they'd have access to this data, and while they're traveling these 16 or 20 hours out there, they could sit there and be evaluating things and coming up with the game plan. And I was really happy to see when they arrived in Guam, they were able to hit the ground running, and they were there for about three or four days, and they were actually able to classify the fire as accidental.

Rod Ammon: Boy, that's a beautiful story. I'm sure you made a lot of people happy at the budget office, but I think it also looks good, I think, when the proper scale of response shows up to a location. I mean, just the thought of... and relationships are important, but just the thought of you showing up with whatever size your NRT would've been in that case or international response. Wow, that's a great story. So, I want to go back just a step and focus on some more tactical, practical level things for fire investigators, because I think everybody's learning, and some of these things are very important, and the details of them. So what types of systems should the investigator be looking for? I know we've touched on it already.

Dawn Dodsworth: Well, the types of systems, obviously, are the ones that everybody is familiar with, which includes your fire protection systems, which would include your sprinkler system, fire alarm systems, burglar alarm systems. And you also want to look for Wi-Fi. Is there a Wi-Fi system in the building? Are there different antennas where you can connect to a Wi-Fi system throughout the structure? We had a case study in Texas where our experts actually were able to map out the different Wi-Fi satellites, and they were able to narrow down the area of origin just based on the failure of those connection points within the building.

Rod Ammon: Nice.

Dawn Dodsworth: You also want to consider, when we're talking about potential cause hypothesis, you might want to consider are there Wi-Fi connection points outside of your building of origin, let's say on a public street where you have public Wi-Fi access. People will be walking, and their phones are going to be connecting to these different Wi-Fi access points. So you may be able to put different people in a specific area during a specific timeframe. It could be just a plain old witness that you might want to track down and interview, because they were in that area around about the time an incident occurred, or it could even potentially be your suspect.

Rod Ammon: Yeah. Wow. I hadn't thought about that side of it. So, the investigators there at the scene, tell me some specific things they might need to ask about these systems.

Dawn Dodsworth: Well, they're going to want to ask, obviously, what types of systems are in the building, and you may have to ask a number of different people, because I've been to some incidents where the manager might know of a couple of the systems, but might not know all of the systems. So I would always use that as a question for your witnesses during your interviews, is what types of systems are in the building that might be connected to Ethernet or might be connected to satellite or might be connected through Wi-Fi? Are there any types of monitored systems that might be monitored remotely by a security company or whatnot? And just try to get as large of a list as possible.

And it's kind of amazing, because we've gone to talking to somebody where they might just have two systems identified to. You speak to a couple other employees, and they throw out another couple systems, so now we're from two systems that are available to obtain data points to five systems. So, you really kind of want to ask that question.

Rod Ammon: Yeah, I think that point about reaching out to different people is a really good one. Oh, go ahead. I'm sorry.

Dawn Dodsworth: Yeah, no, because you're essentially hedging your bets, right? And at the same time, when you find out what systems are there, obviously, you want to consider what type of motion or what type of interaction from the environment might cause that system to activate? For example, there might be door or window alarms, not necessarily from somebody breaking in, but from one of the devices or sensors failing due to heat exposure. So you want to kind of consider, "Okay, what type of ways would I be receiving this data point from this specific system and this specific device or sensor?" Because, once again, the big takeaway here is you have to have proper context when you're using this information.

And that goes right into understanding how something can fail within the building, and again, what type of alarm or what kind of signal would I be getting a trouble signal or a supervisory signal? And the way you could actually kind of... if you're not too certain about that, contact the vendor that may have installed the system and speak to them directly. Also, with a lot of these systems, you're not necessarily going to want to access that data yourself. You might get certain data from a witness like a owner or a manager or whatnot. They may have email notifications of different events taking place, but you're also going to want to get that information off of the control panel. And for something like that, you're really going to need a technical expert that is familiar with that system, because if you try to obtain that information off that panel yourself, you may either inadvertently create new data points that are inaccurate, or you actually may erase, inadvertently, data points that you'll never get back. And then-

Rod Ammon: That's the worst part, huh?

Dawn Dodsworth: Oh yeah. Well, and then the other thing you want to consider is, some of these devices and sensors may actually retain information on it remotely within each individual device and sensor. So, it really helps to contact an engineer. You could contact your local ATF CFI who can get in touch with one of our engineers, or you contact the company that installed the system, and they'll be able to let you know, and they'll be able to send a technician out. Another thing you have to remember is this information is... It may very well be subjected to the Fourth Amendment protection for search and seizure. So, a lot of this information you're going to need either consent, or you may need a subpoena or a search warrant to get.

Rod Ammon: We were moving along so fast, I forgot all about that. I'm sure other people do too.

Dawn Dodsworth: Sorry.

Rod Ammon: No, that's okay. That's a very important point, and think about it often, but as you're cruising through all this opportunity for new information, it's good to remember to check yourself up on those basic things. So, you made one really good point. Well, you've made a lot of good points.

Dawn Dodsworth: I was going to say, "Just one?"

Rod Ammon: Yeah. Well, specifically related to this last topic. Contacting the vendor or the manufacturer of something I think is so powerful. I know that we've had people reach out to us as developers of some things related to CFITrainer, one of the networks that we do, and the amount of information that's available that might not even be available to an administrator of an account often boggles my mind, simple things like logging, which people don't know, that happens to be turned on on a server that's saving gigabytes of text that relates to every touch and click and action. I just thought that was a really good point that you made, and I'd love to see people follow up with that.

Dawn Dodsworth: No, I appreciate that.

Rod Ammon: So, you've touched on some of this, but I'm going to bring it up again, because this is so important to everybody that's out there, about how the investigator can use this data in the origin cause specific determination.

Dawn Dodsworth: As an investigator, what you're going to want to do is, once we go through all these different steps of identifying the system, opining about what ways these different devices and sensors are going to create a data point for you, understanding the system installation and the layout and locations of everything, once you get all that contextual data, and it's normalized, then what you can do is you can sit down with either your investigative team, if there's more than one of you, chances are, if it's a large scene, there is going to be more than one of you, and take a map or a schematic of the structure and... Now remember, we're getting this information of where these different devices and sensors may be located, but then what you want to do is you want to kind of drop down that information into that schematic that you already have. So, essentially, you're going to be mapping out these different data points that you're receiving with the different times and then tracking that information back to your earliest data point that you have within the context.

Rod Ammon: Got it. Okay. Tell us about the strengths of this Internet of Things data, and then, we can go on, obviously, to the weaknesses.

Dawn Dodsworth: There's a number of strengths, and some of those we had gone over previously, which is that it's empirical, and it kind of helps remove some of the subjectivity that might potentially be there. It assists with the accuracy and quality and integrity of the different data points. It's timestamp data that produces a hard time data point for the investigator to consider. It's reliable. Once again, though, it basically comes down to the context of that information. You want to make sure that you are able to put that information within its proper context of how is the system laid out, and is the time normalized? Creates a paper trail, essentially for the investigator to look at and trace back. And it really kind of provides more of a black-and-white empirical picture, and it's more objective in nature, as opposed to being subjective. And when we look at the 921, this new recent iteration, the arc mapping was moved, essentially, under fire patterns. Well, this itself also creates, almost essentially, a fire pattern, because you're actually tracking the pattern of device and sensor activations and failures within a building, and you're looking at that in context with the entire system and within the context of the building.

Rod Ammon: Interesting way to put it. So, thanks for going over those again, because I really think that... Again, I love getting some of these more tactical things down to people, and I know Cathy, who does our writing and things, would be proud that I'm getting some of these things repeated, because a lot of what we're trying to do is educate. So weaknesses.

Dawn Dodsworth: With anything else, there's always going to be some potential issues or weaknesses within a technique, and that's why we utilize a bunch of different tools in our toolbox, and we don't rely specifically on one particular tool. So the biggest thing, really, again, and I can't say it enough, is you need to have the proper context. As an investigator, you need to understand the layout of the system, the type of the system, what the construction is of the system, the devices and sensors. Where are the locations of them? How is the system installed? And you also need to make sure that the times are normalized, because if you have a lack of context, obviously, you're going to then have a poor interpretation of that data, and the potential is there for it to be improperly applied to the scene, and then it may lead to misinformation that will not aid you and narrow down the area of origin or the fire progression.

There's the potential for human error. Again, like I said, inaccurate application, some type of error of commission. You may incidentally... Let's say you don't reach out and contact the technician, and you might accidentally omit or delete information. So, you just have to kind of be cognizant of the fact that that potential is there and work around it.

Rod Ammon: Yeah, I get it. I guess that leads to something we dealt with back in the day when we were working on some things related to electronic data with Secret Service, and I know they were experts. Can you talk about people that an investigator might be able to reach out to for expertise?

Dawn Dodsworth: Sure. And that kind of comes in towards the conclusion of my research. I kind of came up with a number of areas that really need to be researched further to try to make it even easier for investigators. And one of those was developing a system of best practices, not for each individual business and each individual system for specific businesses, but kind of like a general best practices of these different types of systems. And one of that would include like, "Hey, who do I need to reach out to get this information?" Because sometimes, the data comes out, and it's plain as day, and it's just in regular verbiage that an investigator can understand. And then sometimes, the data points may come out in some type of a code that then has to be interpreted by an engineer or somebody familiar with that specific system, some type of technician. So, you have to keep that in mind. And then also, occasionally, you're going to have to reach out to not only the manufacturer, but you might have to reach out to the company that actually has come up with the particular design of each system and get information from them. And then, sometimes, the issues with that is they may be hesitant to provide you with that information, not only because of what is one of big issues we have nowadays, liability concerns, and the other is it may be data that is protected. It's something that they may have the market cornered on, and they don't want to share that data. So, again, you have to remember that it's going to be subjected to Fourth Amendment. You might have consent from the owner, but you might have to get a subpoena or a search warrant to get any information from the company if they choose to not be as helpful as they should be.

Rod Ammon: Got it. Yeah, a lot to remember. So I'm thinking learning from case studies, and you started to talk about one, and you teased us a little bit at the beginning. You want to give us some examples of how you've seen the Internet of Things make a difference?

Dawn Dodsworth: Oh, sure, absolutely. I went through a bunch of different cases. There were some really interesting ones out there that are still going through the court system, so unfortunately, I was not able to use those. But one goes back to a fire back in 2005 at a wine storage warehouse in California. So the building size was about 240,000 square feet, and it was about a $450 million loss. And this building inside, they just contained storage of wine from all over the world, and part of the building had a mezzanine on it. And what our agents were able to do was go in there and map out the beam detectors, smoke detectors within that building, and they were actually able to map out where the activations took place, and when they mapped out the activations that took place, it put them specifically right in front of this one cage owned by a particular gentleman on the mezzanine. And then what they did was they began excavating that area, and lo and behold, they ended up finding a propane torch that was laying on the ground there by a bunch of cardboard and was actually still on the on position. So as a result of that, they were actually able to classify the fire as incendiary, and basically, the person that set the fire was involved with essentially some type of like Ponzi scheme with the selling and purchasing of wine, and he would sell wine, but he wouldn't pay the people back that he was storing the wine for. And then he was using that money for other things, and then, essentially, when his clients were starting to close in on him, wondering where their money was, this fire happens. So the defendant ended up getting 27 years in prison for that particular fire.

Rod Ammon: Wow. So I need clarity on one thing. You said the propane torch was in the on position?

Dawn Dodsworth: Yeah, they had looked at it, and it actually so happens that my husband was... He was the lead CFI on that fire.

Rod Ammon: And so, what was the idea that this person left it on and walked away?

Dawn Dodsworth: Yes, yep, give themselves time, get out of the building and have a little bit of a delay between the fire starting. And you have all this available fuel, cardboard and everything else in the area, that's going to ignite fairly readily, and just walked away and left.

Rod Ammon: Maybe I'm going to sound naive, but I've been around you guys for almost 20 years, and I don't think I've heard about somebody... I've heard people leaving gas on, but not like a propane torch, and it just seems like it's sort of like a new way of leaving the gas can behind. Does this happen often?

Dawn Dodsworth: There's just no end to the ingenuity of somebody that really wants to do something. They make it happen.

Rod Ammon: It just seems like... I just can't imagine you'd expect that this tank in its on position would burn up, but okay.

Dawn Dodsworth: Well, what it was, is essentially it was one of those little torches that you might use to light your firepit or whatnot, and it has that little screw knob on it, and then you just click it, and the gas comes out and it forms a flame and that was just left laying there.

Rod Ammon: Wow. Good story. I think you have another.

Dawn Dodsworth: Yeah, so in Kansas City, Missouri, back in 2015 there was a fire that ultimately resulted in the death of two firefighters. It was a mixed commercial and residential structure. Overall size of the building was probably about 17,000 square feet. So our National Response Team got called in to assist, obviously, because I mean, it's going to be very taxing on the locals, and there's firefighter deaths, and we want to make sure that a very thorough job is done and all our experts come in.

So altogether, the building was 14 commercial, or I'm sorry, four commercial occupancies on the first floor and 16 apartments on the two stories above. And while we were in the process of processing the scene, and obviously, you're doing a neighborhood canvas, and you're looking for... The new thing nowadays that kind of makes a lot of people's lives easier is having something on video, so they were canvasing looking for video and talking to neighbors and other businesses, and just so happens that when they went around the back of the structure, there was a bank of smart meters. So what these meters do, they're the electric meter. They record the electrical usage within units of a structure, and what they do is they collect data on the power consumption in 15-minute intervals, and then every four hours, for this particular system, it ends up transmitting this data. So these meters have an internal memory, and then that data, it gets transmitted to the company so that they can have a reading where they're located. So what our electrical engineer did, in conjunction with our lead CFI, was essentially, start to create a pattern of some type of a background usage for essentially a baseline of what was the general usage of electricity in that building in those units for about two weeks prior to the incident. And this way, once again, it gives you the context when you're looking at different things that are going on. You have some kind of solid foundation to build off of. And what they did was they looked at not only the house power for the structure, but then the usage for the individual units. And on the day of the fire, they were kind of looking at when the power fell to zero.

So by doing this, they were able to determine that the unit of origin was a nails and spa salon that was located on the first floor, because trying to time the loss of the electrical power for where these different conductors are running through the building, so everything was mapped out. And then they actually were able to have some video of the suspect leaving the scene. They also did further research and were able to determine that this person was involved in a couple other insurance fraud schemes involving fire. So ultimately, the defendant was charged with two counts of murder and arson charges. The fire was classified, obviously, as an incendiary fire, and the defendant received a 74-year prison sentence.

Rod Ammon: Wow. So help me out again, because I may have missed it while I was sitting here looking at notes. I'm trying to... Where was it that the identity... that you were able to make the determination of incendiary?

Dawn Dodsworth: So what they essentially did was they were able to narrow down the area of origin in this building to the northeast corner of the first floor storage room within that nail salon and spa. And then they were able to go in there and look at any potential accidental causes that could have occurred, and they were able to eliminate those. Then at the same time, they looked at the video of when this person was seen leaving to when we start getting the first traces of an incident taking place, and then, combining that with the contextual data from the smart meters to build out a timeline.

Rod Ammon: Okay. Thank you for the clarity.

Dawn Dodsworth: Sure.

Rod Ammon: Even if it was only for me.

Dawn Dodsworth: Yeah.

Rod Ammon: I am so grateful. I'm wondering what we're missing. You've been wonderfully concise. I think the information that you're involved in is great. I love the fact that you did your dissertation on this, but I want to make sure we're giving you an opportunity to wrap up, or is there anything that I missed?

Dawn Dodsworth: I think the biggest thing is there's still a lot of research that needs to be done. Luckily, we have some new CFI candidates for the ATF program going through the program where they have to do their own individual research projects. So I'm going to probably reach out to a couple of them and see if anybody kind of wants to work with me with pushing this research forward. What I would really like to do, and I know we have some engineers that are working on it now, is this overall large concept of these big fire scenes that I used as case studies, it really would be great to narrow that down to smaller structures such as residences and looking at maybe smart appliances and smart homes and Ring camera systems and Nest thermostats and how those systems and devices can be utilized, maybe, to narrow down a timeframe or an area of origin within a residence, maybe looking at the temperature reading within a building or looking at when different activations or signals come through, or maybe even looking at humidity levels within the building. And then furthermore, applying this to vehicles, looking at maybe a vehicle fire. A lot of your newer vehicles nowadays have what are called infotainment systems, and those infotainment systems can be downloaded, and they provide a plethora of information. I mean, it could be from when were the doors locked, or when were the doors unlocked? When was the door opened? How long were you driving for? What speed you were driving. And then the amazing thing, also, is, if you have your cell phone in these vehicles, and a lot of people like to connect their phone through Bluetooth in the vehicle, well, the potential is there that that entire phone could have been downloaded into the vehicle memory system, and now the investigator not only has the information on the vehicle operations and whatnot, but they also have information directly off of the person's phone.

Rod Ammon: Yeah. It's amazing how much information's captured and how much information is sent from the vehicle. I know Toyota, I think they're tracking my every move on a new vehicle, and I just think, "Wow." They write me messages that are obviously specific to my mileage. They ask me... The speed issue is interesting. And I know with insurance companies, they're talking about looking at people's driver habits, and they're watching acceleration, deceleration, all those things. Well, I think it used to be a little easier to get away.

Dawn Dodsworth: Yeah. I mean, it's almost like Big Brother's watching everywhere now.

Rod Ammon: Yeah. Well, I guess for us that are following the law, it's not as much of a concern, but-

Dawn Dodsworth: Very true.

Rod Ammon: I appreciate your time today so much, and this is interesting to me personally, and timely, I think, especially as all of this technology is so rapidly expanding. Wow. It's just like I said earlier, a lot more to keep up with for fire investigators, but I think, at the same time, that learning about that is also motivating, or it would be to me. I can imagine it would be. So, thank you-

Dawn Dodsworth: Yes-

Rod Ammon: ... for giving everybody this time.

Dawn Dodsworth: Absolutely. I appreciate you having me.

Rod Ammon: Well, we're very grateful. And we'll look forward to seeing you in the future.

Dawn Dodsworth: Absolutely.

Rod Ammon: Something tells me we will. Thanks again, Dawn.

Dawn Dodsworth: Thanks, Rod. Thank you.

Rod Ammon: Bye-bye.

Dawn Dodsworth: Take care.

Rod Ammon: You too. Now for some news from the IAAI, two scholarships are currently open to members of the IAAI until October 1st. That's coming up pretty quick here, so take advantage of this opportunity. There is one $5,000 Foundation Academic Scholarship that will be awarded to provide financial assistance to students pursuing formal education in a curriculum consistent with the mission of the IAAI. There are also five $1,000 Foundation Training Scholarships that are awarded to selected applicants to enhance their education in the fire explosion investigation profession by attending the IAAI International Training Conference, or other select IAAI training events. Go to firearson.com or iaaifoundation.net for more information and apply today. The deadline is Saturday, October 1st. This is our 100th episode of the CFITrainer.Net Podcast. We'd like to take a moment to thank you for your support. Thousands of you tune in every month, and we're grateful for the many messages you send us telling how you've enjoyed the episodes and the value that they've had in your work. It's your support that keeps us going and pushes us to find new topics that will add value to the investigations you participate in. Thank you, and thank you for your work that you're doing out there in the field every day to prevent fires, care for victims, and hold criminals accountable for their actions. Please let us know what you're thinking. Your opinions and suggestions are really appreciated. You can use the feedback form on this page.

This podcast and CFITrainer.Net are made possible by funding from the Fire Prevention and Safety Grant from the Assistance to Firefighters Grant Program administered by FEMA and the US Department of Homeland Security. We also get support from the Bureau of Alcohol, Tobacco, Firearms and Explosives and voluntary online donations from CFITrainer.Net users and podcast listeners. Thanks for joining us today on the podcast. Stay safe, and we'll see you next month. For the International Association of Arson Investigators and CFITrainer.Net, I'm Rod Ammon.

Dodsworth, Dawn. Internet of Things (IoT) Connected Building Systems and Their Use in Fire Investigations: A Mixed Method Study. Doctoral Dissertation. Oklahoma State University. December 2021.

NFPA 1030: Standard for Professional Qualifications for Fire Prevention Program Positions. National Fire Protection Association