Episode 330: Building for the City of the Future: CAD Data, Digital Twins and Station Placement with Craig Hiltz – Darkhorse

Episode notes

Every fire department already owns years of CAD and RMS data. Most of it sits untouched…

Hello Smart Firefighting Community!

Welcome to another episode of covering real world innovations via interviews with fire service and technology industry experts that empower YOU to develop your very own Smart Firefighting strategy.

Craig Hiltz of Darkhorse turns that data into a digital twin of the response system, so a chief can test a station, a roving engine or a staffing change on screen before spending a dollar on it.

Recorded at Interschutz , this conversation walks through what a digital twin is in plain language, why a dashboard only tells you what happened yesterday, and what departments in Meridian, Odessa and Toronto found when they asked the what-if questions. Craig also shares his own path from statistician and market researcher to the fire service.

Topics covered:

Why CAD and RMS data goes unused and what it can answer

Dashboards versus the why and the what if

Digital twins explained in plain language

Meridian, Boise and Eagle in one regional model

A roving engine instead of a new station

Accreditation data packages: SOC, CRA and the three axes risk model

Odessa: two minutes off dispatch equals ten new stations

Toronto: a ten percent budget cut and NFPA 1710 compliance

ALS to BLS, fleet decisions and firefighter call stress

Craig's path from statistics and market research to Dark Horse

So take a listen to hear what he has to share.

Head to www.smartfirefighting.com to discover how SFF accelerates innovation for emergency responders, to find out when our next event is, or review our curated resources.

Facebook | Instagram | Twitter | LinkedIn

Chapters

1. 0:00 Interschutz mini-series intro
2. 0:24 Craig's first Interschutz
3. 1:24 What Darkhorse does: unlocking CAD and RMS data
4. 3:03 Planning for growth years out
5. 3:36 Meridian: accreditation and where to build the next station
6. 4:46 A roving engine instead of a new station
7. 5:28 Accreditation data packages: SOC, CRA, and the three axes model
8. 6:25 Risk first, then response
9. 7:02 Are dashboards dead?
10. 8:22 Digital twins explained simply
11. 10:03 What you can simulate: stations, staffing, and units
12. 13:01 Odessa, Texas and dispatch times
13. 14:21 Toronto and NFPA 1710
14. 16:10 Craig's path from statistician to Darkhorse
15. 17:42 Words to live by

Questions answered in this episode

What is Dark Horse and what does it do for fire departments?

Craig Hiltz says Dark Horse takes data that departments already have in their RMS and CAD systems and builds a simulated environment for the response environment. This lets departments make changes within their deployment region and see what it would do to their response times, utilization rates, and fleet management.

Hear it at 1:22

What is a digital twin in the context of fire service operations?

Craig Hiltz explains a digital twin as capturing all the behaviors within a particular environment so that a screen on your computer behaves the same way as real life. If you add a new truck to a system, the digital twin shows you the same gain in performance before you actually add it, acting as a risk mitigation environment to test ideas.

Hear it at 8:44

How did Dark Horse help Meridian Fire Department with station placement?

Craig Hiltz says Chief Blume's team in Meridian was going to build a station in a new area. Through working through scenarios, they determined they should put a roving engine around to fill in the stop gaps. They got more improvement to their ERF by that in the immediate future than building that station, so they could defer it.

Hear it at 4:09

What did Dark Horse discover about dispatch time in Odessa, Texas?

Craig Hiltz says they helped Odessa understand that reducing their dispatch time by two minutes was equivalent to adding ten new stations in terms of first due arrival performance gain.

Hear it at 13:17

What is the three axes model that Dark Horse developed?

Craig Hiltz describes the three axes model as looking at not just the consequence and probability of an event, but also the impact of that event. It can show geospatially what part of the city has the highest risks and identify whether impact or consequence is the biggest component of risk for different groups of incidents.

Hear it at 5:48

Transcript

Machine transcript of the published audio. Names may be misspelled and some sections are missing.

[0:20] Kevin Sofen: On the tail end of the Interschutz experience. What's going on, Craig? Good to see you again.

[0:25] Craig Hiltz: You as well.

[0:26] Kevin Sofen: How is your Interschutz experience treating you?

[0:29] Craig Hiltz: This is our first time to Interschutz and it's quite the experience. I've been to a lot of conferences in North America and Australia, but this is definitely the largest trade show I've ever been to.

[0:40] Kevin Sofen: It is the largest trade show conference exposition period in the world. Inside, outside, the number of trucks here is amazing and stuff that I didn't even know existed in the fire industry.

[0:51] Craig Hiltz: It's still head-scratching. I don't understand how there are going to be that many pump companies, that many fire extinguisher companies, that many valve companies, but I guess there are 8 billion people in the world. And all the clothing, the gear is amazing how much there is.

[1:08] Kevin Sofen: I'm sure it's needed.

[1:09] Craig Hiltz: It is.

[1:09] Kevin Sofen: So, data, fire service, making better decisions, Dark Horse, a couple words that I kind of maybe define. Come to mind. They come to mind when I think of Craig. Dark Horse. What is Dark Horse and give me the high-level scoop.

[1:22] Craig Hiltz: Essentially, Dark Horse is taking data that the department already has in the RMS and CAD systems that they've had for years, but haven't extracted the full value out of it. So, we're taking that and putting it into a, building a simulated environment for the response environment so that they can actually make changes within their deployment region and see what it would do to their response times, their utilization rates, their, in some cases, fleet management and other things. So, it's really just unlocking the value that they already have within the department but haven't done it yet.

[1:57] Kevin Sofen: So, they already have this data? Like, where is this data and what format is it now?

[2:02] Craig Hiltz: It comes to know their CAD and their RMS systems. So, it's sitting there.

[2:06] Kevin Sofen: But it's in what, like, just gibberish ones and zeros?

[2:10] Craig Hiltz: Well, they have reports within their RMS and CAD systems, but they're what we would call a dashboard. So, a dashboard is telling you what. Then there's the why. Why is it happening? And then there's the what if. What if I change something? And the why side and the what if sides are a bit more difficult to tease out of the current tools.

[2:29] Kevin Sofen: The what if, I would argue, they can't.

[2:32] Craig Hiltz: Right? And so, that's where this digital twin concept comes in and we're able to really make the data transparent in a way that those that maybe aren't as comfortable with data cannot actually see what is happening and gain trust in that. Because it's all about gaining the trust of those that you need to have on side to make the big decisions. And we're talking long-term strategic decisions. We deal more in what's going to happen next year, five years, 25 years out. So, we'll get planning information from cities, city planning departments, saying in 2025, we're going to have two million more people. That is going to be X million dollars or billion dollars in industrial development here, here, and here. And that warehouse district is going to be high-rise, it's filled with old people. What's your future demand going to be? Well, that's what you're building for. You're building for the city of the future, not for today. So, that's where we help the most.

[3:28] Kevin Sofen: So, building for the city of the future, I like that. Because I heard Chief Blume from Meridian say something similar to that. And he pointed me when he saw you, he was like, "Man, Dark Horse helped me with my accreditation process and also helped me figure out where we were going to build our next fire station." And I was like, "Okay, well, tell me more." And I'm curious from your perspective about, again, taking all this historical data that's sitting in some obscure CAD files and then using that to enable departments to, you know, A, help both with accreditation, but B, also confidently say, "Not only do I know I need the fire department, but this is exactly where I'm going to build it."

[4:09] Craig Hiltz: And, I mean, they got a really smart team there to start with. So, they came from a really good spot, but they didn't have the capabilities to do this kind of stuff we could help them with. So, they, yes, got accredited. And interesting, once the Meridian department came on board, then Boise came on board and Eagle and the whole region. Because it's really interesting when you have the region all in the same tool, because then you can understand the borderless system and where resources should all be cooperating, which is kind of the new world. Chief Blume and his team, they had a whole new area they were building. They were going to build a station out there. Through working through some of the scenarios, they determined that maybe they should have a, I think they put a roving engine around to kind of fill us on the stop gaps. And they got more improvement to their ERF by that in the immediate future than building that station. So, they could defer that, I don't know how much their stations are in Meridian, but non-trivial amount of money. They still need it, but they were going to use it as a stop gap for some other issues. And now they've been able to defer it a bit, I think is still the plan last time I talked to them. And still gain the improvements they needed to you. And then the accreditation stuff is, again, there's a bit of a process to get accredited. And so, we have all the data they need, both on the risk and the response side. So, the SOC and the CRA side, to be able to allow them just to put a pretty comprehensive data package together with not a lot of work. So, we have the three axes model, is the newest thing we came up with. And it's...

[5:48] Kevin Sofen: What's that?

[5:48] Craig Hiltz: Three axes model is when you're looking at, you know, not just the consequence and the probability of an event, but also the impact of that event. And so, we can do a two axes model geospatially that shows you what part of the city you're having highest risks in. But when you bring in the third axes, the impact, now you're starting to look at, for my different groups of incidents, and my medium risk fire incidents, what is the biggest cause, or what's the biggest component of a risk in those. So, sometimes it's impact, sometimes it's consequence. These are the medical, it's more consequence. And that's really what the CPSC group is needing to provide the accreditation piece. So, it's important. And it's really the risks should be looked at first, and then the response. We kind of have it backwards in general, because we're looking at our SOC first, and then risk is sometimes an afterthought.

[6:41] Kevin Sofen: And earlier you said the word dashboard. And sometimes dashboards obviously are great, but I also heard the phrase at one point that dashboards are dead. I don't know if that's completely true, but I do think that sometimes, just when you have a... You may be able to convert some ones and zeros into a nice dashboard, there's still that level of like, comprehension of what does the data mean. You know, understanding where the data came from, but then also not only what it means, like you said, of like, well, what can we do with this data now? I have a nice PDF or a PNG of some graphs, of some charts, whatever. But I'm curious in that your take on dashboards are dead, agree or disagree, and then kind of how you can, how you help provide context to the before, during, and after in the dashboard life.

[7:29] Craig Hiltz: Yeah, I mean, I don't know if dashboards are dead per se. I think dashboards alone aren't enough to tell the story. A dashboard is a quick reference to see if we're getting better or worse today versus some other time. But the real question is, if you're not getting better, why aren't you getting better? And the dashboard has limited capability to help you with that. So I think the dashboard is kind of like your newspaper. You know, it tells you what happened yesterday, but it doesn't tell you what's going to happen in the future. So I think in combination, we still have dashboards. I mean, they're still useful. You're trying to pump them out in a little more real time and kind of a little more automated. But feel like the dashboard alone is not enough.

[8:13] Kevin Sofen: I like that. It's like a newspaper because, yeah, it's good to know what was in the news. But I think a lot of times you talk to first responders, they're obviously kind of thinking about planning for tomorrow. And so when you're saying it's not enough, and then you're thinking about layering on some of this predictive analytics and digital twin, I think some people may hear that, and that's kind of over their head. Can you explain digital twin and predictive analytics in a way that maybe a third grader could understand?

[8:44] Craig Hiltz: [inaudible 8:44] try. I mean, the digital twin is essentially capturing all the behaviors within a particular environment so that that screen on your computer is going to behave the same way that would happen when you're out in the real life. So if you were to add a new truck to a system, your city, you'd expect to get a gain in performance. Your digital twin, if it's good, would show you that same gain in performance before you actually add it to the editor. So it's basically a risk mitigation environment. Say, I can test stuff and it comes from my own office. I don't have to risk getting, [inaudible 9:18] mean, rightfully so, not good feedback from the front line when it doesn't go well because I've already got rid of those 90% of the kind of bad ideas. I have them, you have them. It's a great idea when you come up with it. When you actually put it to the test, it doesn't kind of meet that litmus test, and you're saying, oh, that was great that night, but when I thought about it tomorrow, it wasn't a good idea. So this idea about digital twin, it's a kind of filter those out. And in a lot of our clients, provides new ideas that they never thought about. We have a famous client who said, that was the 11th idea I had. We had 10, but our analysts gave them an 11th, and that's when they went with, because it constructed through understanding the why tool, why is something happening, and I want this outcome. If you take those two goal posts, you find out what that path is to get to that outcome.

[10:11] Kevin Sofen: And I want to try and get a little more granular in terms of that, maybe the 10th and the 11th idea of, are we talking about regards to, I know you mentioned fire station placement to help understand response time, to then, in my head, I'm also thinking about simulating evacuation routes to, again, staffing. What are some of these things that you can start to simulate and play around with once you have a digital twin?

[10:35] Craig Hiltz: Yeah, I mean, we've had certain jurisdictions where COVID funding comes off and their ambulance system isn't fully funded. We might have a difficulty getting paramedics, so they need to move from maybe an ALS to a BLS system. Kind of a controversial move in some places, but we can help them understand, if you make that move, here's going to be the corresponding change in performance, and by through that, you can communicate that to your political people, union partners, and the rest of it. So that's one. Understanding fleet management, it's like you're running the wheels off your engine when you could have 350 or a 450 doing the same thing, F450 doing the same job, right? So, you know, engines are not cheap, and coming up with other ideas where I could swap out apparatus and still, you know, deliver the services I need, that's a real thing. Or we're looking now at things like we can understand how many calls a particular firefighter has been to that are, you know, higher stress than usual, and, you know, maybe that can be provided back to the captain to understand what's going on. So not so much predictive, but it's about using the data we have and answering the challenges that exist in the fire department. So interesting. Our partners usually come in and they clean up their own backyard, they fix the dispatch system, they do whatever that is. They do the long-term planning, which is look out 25, 15 years, understand what that kind of build out will look like for their strategic plan, inform the city councils or their boards. And then they do kind of a short-term resource optimization of this is what I have to work with right now, how can I get the best out of it? And after those two are done, it's kind of, gloves are off and they come up with some of the most interesting questions, because by that time they become comfortable with and trust the analytics, the predictions, and now they're answering some of these more unique questions that our fire analysts will step in and help out with. And stuff you never even thought of, but now that they have the ability to make context and make sense of it, they can start to push limits. Quite honestly, things we have to think about. It's like, that's a great question. And so you kind of go back and some of our best inventions or additions to the platform have been coming from a chief asking a question. It's like, never thought about it, but I think we can do it.

[12:58] Kevin Sofen: Yeah, and I know I brought up Chief Blume earlier and he was very, he told himself to be part of a potential conversation. But are there any other one or two customer case studies that you can share that are public that might kind of paint a picture on again how Dark Horse can be kind of an engine for good for others?

[13:17] Craig Hiltz: I mean, Odessa, Odessa, Texas is doing some really, really cool stuff. I mean, down there they had a long dispatch time, and I won't get into why, I think most people know, but we help them understand that if they could reduce their dispatch time by two minutes by changing and deploying a bit earlier, it was equivalent to them adding ten new stations. So they would have to add ten new stations to get the same first-due arrival performance gain as they did by chopping two minutes off their dispatch. So the idea is like, what's that trade-off? What's it worth to drop that dispatch? So you might have to pay for more gas because you're calling them off more often, but is that equivalent to ten million ten stations, like whatever that's going to work out to you in Odessa? So that was a really cool case study, and it really comes down to what's the trade-off. I mean, often we're deemed as a cost to a department, but we try to view ourselves as being how can we help them become more efficient, which is going to either give them money to spending other areas that are so often needed within a department or maybe not have to add so many operational resources as their population grows. So that's one. And Toronto, I mean, we went into Toronto, it's like ten years ago, but they had had a ten percent budget cut soon after we got there, and it was like one afternoon. And so we helped them understand where to kind of reduce their budget cuts, and then over the next, I think, 18 months or 12 months through kind of focusing on the operations, they were back up. They were actually the first NFPA 1710 meeting that criteria in North America at that point ten years ago. And so it was like with the budget cut, they still exceeded or improved and made some very aggressive performance metrics.

[15:02] Kevin Sofen: That's the epitome of the fire service, how to do more with less.

[15:06] Craig Hiltz: Yeah, it is. And unfortunately it's true, and it's only going to get worse. And then fire is a whole different story, right? We do a lot on risk, community risk, so it's how can you, and the goal there is to reduce distractions. How to reduce-- How to reduce the damage. How to reduce the damage, the whistle. How is that going? Yeah, what were we talking about? Yeah, how to reduce number of costs through prevention. And so like Chief Blume's groups are doing great jobs there too. They're using some of the tools and having some real success on it. So not just the software, it has to have the culture internally. Yeah. And what we try to do is support that culture so it's not a technical hurdle. So people internally that really have the vision, have the tools to be able to get there.

[15:52] Kevin Sofen: So Craig, I'm curious, obviously I've known you for a few years now, but I never really knew Craig before Dark Horse. How did you get into Dark Horse? What would you do just being this, as I mentioned, data efficiency, fire service, data wizard guru guy? Yeah. How did you get into this?

[16:09] Craig Hiltz: My life before, I'm a statistician by training. I got a graduate degree in stats. I used to sit there doing the models in a dark room. Nobody used to talk to me. You know, they'd lock the door. Don't look at me. Don't talk to me. Don't talk to me. And I really despise those guys that went out and promised everything to clients, which now is me. But, you know, I grew up and eventually we had a market research company that worked with the likes of Apple and HP and, you know, Spin Toothbrush. We developed that Voice of Internet Protocol in Canada. We did that. And so this is really HP printer. You can blame me for the HP printer, the ink on it so much. That was our idea. But the point was that we, you know, I grew up that environment, but I was always dealing with kind of new age technology, which people had never seen before. And my job was to measure how they should put those products together so that consumers would see value in them. And, you know, I spent 20 years on there. We built a great company. I think we went from six people to about 500 when I left. Wow. And so I, said bye to my partners and kind of took a couple of years off. And then I found Dan Haight, who is leading this company. And I thought it just had a very strong technical background. He had some really cool folks there, really smart, really committed. And that's how I got involved. So it's, you know, the fire industry has been very kind to me. I'm not a firefighter, but they've adopted me. And, you know, I try to not pretend I'm a firefighter. I bring what I bring, and I think it contributes to the environment. So it's been fun. It's been really fun and quite rewarding, actually.

[17:47] Kevin Sofen: Yeah. Likewise, Craig. Well, to wrap this up here, I'd love to know within your life as a statistician or your life within Dark Horse and emerging in the two, do you have any favorite quote or words of wisdom or kind of phrase that you think about or tell your team or that you live by?

[18:04] Craig Hiltz: Well, I mean, the one I used to tell my team last time was in God, we live in God, everybody else must bring data, right? Like, you know, you can believe something, but I'm a big fan of, you know, show it to me, prove it to me. And, you know, that's kind of what we have a foundation. You know, there's no room for, I believe this is happening. My gut says this. My gut says this. I mean, that's the best direction to go, but it's got to be confirmed with something that's a little more defensible. So I'm not a big coach guy. I apologize.

[18:36] Kevin Sofen: That was a mic drop. That way, yeah. [inaudible 18:40] well, Craig, I really admire you and your team. It's been great to see and experience some of the stuff that Dark Horse is doing and look forward to continue to see and monitor all the great stuff that Dark Horse does for public safety. So thank you.

[18:53] Craig Hiltz: Well, thank you for having me. It's been fun. It's been great to be at Interschutz. And the party you had last night was fantastic. It was good. I was pretty happy man. I had pretzels, beer, a roof over my head, music and just networking. So I could basically die a happy man. It was very great. It was very good.

[19:11] Kevin Sofen: Congratulations.

[19:12] Craig Hiltz: Thank you very much. All right.

[19:14] Kevin Sofen: All right, Craig. Take care. Take care.