Smart Firefighting Podcast

Episode 275: Using IoT and AI for Fire Detection with Rob Vernon

Published

Guests: Rob Vernon

Topics: Artificial Intelligence, Data & CAD, Wildfire

Episode notes

Let’s talk about the importance of measuring what matters in the fight against climate change…

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!

In this episode:

Leveraging IoT, computer vision, and AI for wildfire detection

How data aggregation helps first responders with real-time detection and triage

Discover the goals of the XPRIZE wildfire competition and how it aims to revolutionize fire response

Hear from Rob Vernon - the CEO of Indicium Dynamics .

Rob shares insights into their participation in the XPRIZE wildfire competition, focusing on innovative approaches to fire detection and response. Working alongside a team of partners, Indicinium uses IoT, computer vision, and AI to aggregate and analyze data, providing high-quality detection and triage services to first responders. Rob discusses their plans to have operational services using drones for situational awareness by Christmas, with a key milestone set for March.

The XPRIZE competition emphasizes both rapid detection and automated response to fires, aiming to transform how wildfire incidents are managed. Throughout the conversation, Rob stresses the importance of "measuring what matters" in tackling climate change and using data effectively.

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

Episode Resources:

AFAC25 Conference

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 Intro from the IFE booth at AFAC 2024
  2. 0:50 Rob Vernon joins
  3. 1:11 What the XPRIZE wildfire competition is
  4. 2:21 Collaboration over competition
  5. 3:00 Aggregating disparate data sources
  6. 3:47 Multi-vector detection and triage
  7. 5:46 The path forward after XPRIZE
  8. 7:40 It takes a village: Track A and Track B
  9. 8:28 Indicium, Latin for information
  10. 8:38 Tim Tams
  11. 9:04 Measure what matters
  12. 9:43 Wrap

Questions answered in this episode

What is the XPRIZE wildfire competition trying to achieve?

Rob Vernon says the competition is about detecting a fire from space, the size of a booth, under a minute and reporting it back to earth in under 10 minutes. The other side is to go and do something about that fire, essentially put it out.

Hear it at 1:18

How does Indicium Dynamics detect fires?

Rob Vernon explains that detection is multi-vector. They use sensors to smell for smoke, cameras to visually look for smoke, and Earth observation from satellites. When they get a detection from one vector, they try to validate it from another, such as triangulation.

Hear it at 4:19

What service will Indicium offer by Christmas?

Rob Vernon says by Christmas they plan to do terrestrial detections combined with automated drone deployments for situational awareness. They will detect a fire and stick a drone over it with optical and thermal cameras to start mapping the fire boundaries.

Hear it at 6:01

What is the difference between Track A and Track B in the XPRIZE competition?

Track A is the detection from space and communication under 10 minutes. Track B is the automated response, the go and put the fire out piece. Rob Vernon says Indicium is the orchestration engine responsible for terrestrial detections, Earth observation detections, and automated response.

Hear it at 7:51

What is Rob Vernon's guiding principle for the company?

Rob Vernon says the catchphrase is measure what matters. You measure something in order to baseline it and then probably change it, but you have to figure out what matters to you as an organization.

Hear it at 9:06

Transcript

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

Ep 275 transcript

Using IoT and AI for Fire Detection with Rob Vernon

Speakers: Kevin Sofen, Rob Vernon

Source: Whisper medium.en machine transcription of the published Podbean audio (no original transcript found in Drive), lightly cleaned. Speakers were separated by voice clustering and then checked and assigned by hand at every speaker change. Gaps where the model returned no usable text are marked [inaudible m:ss].

[0:00] Kevin Sofen: Welcome to a mini Smart Firefighting podcast series, recorded live at the Institute of Fire Engineers booth during the 2024 AFAC Show Down Under in Sydney, Australia. Throughout these conversations, we dive into critical issues like fire safety, emerging technologies, diversity in the fire service, and innovative solutions for emergency response. with these fascinating insights from leaders around the world on how we can build a safer and more resilient future. Rob Vernon, CTO of Indicium Dynamics, shares our innovative use of IoT and AI in wildfire detection. He discusses their participation in XPRIZE competition, their plans for drone-based fire detection services, and the importance of using data effectively to tackle climate change. All right, all right, sitting alongside Rob Vernon from Indicium Dynamics.

[0:56] Rob Vernon: Everyone gets it on the third go.

[0:58] Kevin Sofen: Okay, by the end of this podcast, I will officially get it. Very good.

[1:02] Rob Vernon: So how you doing? Yeah, well, man, thank you for having me on.

[1:04] Kevin Sofen: Hey, this is good. Well, I'm grateful for Andrea and Colin and Claudia and the XPRIZE team who connect us here today.

[1:11] Rob Vernon: Yeah, they're great.

[1:11] Kevin Sofen: And so we're talking about XPRIZE on a high level. What is XPRIZE for people who have maybe never heard of it?

[1:18] Rob Vernon: So the XPRIZE wildfire competition is about, in essence, trying to detect a fire from space, you know, the size of this booth that we're in under a minute and then report that back down to earth in under 10 minutes. And then the other side of the competition is, and go do something about that. So go and put the fire out in essence. So, yeah, we've, we've formed a team up to, in essence, tackle that problem. Ourselves at Indicium, we're an IoT company. We're doing things with computer vision and detecting smoke using AI. That's one part of the jigsaw puzzle. And fire for us is quite a holistic effort. So the team that we formed up ourselves at Indicium, we've got Little Place Labs out of Texas who are doing the Earth observation, the satellites in space, virtual constellation. The guys out of Hong Kong are robotics cats on some more computer vision stuff. And some guys are getting closer to home for us, TAS Drone Solutions, who are working with us on the autonomous drone response. So yeah, exciting stuff. My pitch to those guys was that the competition's a great North Star for us. There's great business to be had in this space. So let's all put our heads together on that.

[2:21] Kevin Sofen: Yeah, I really appreciate that about XPRIZE of being this community focused on collaboration and working with each other. And it's definitely the epitome of a moonshot style problem. But it is crazy thinking about all these different means of new data from sensors that are smelling for smoke to cameras up in the sky from drones that are passively searching as well as satellites. So there's all these different new means of how we are creating data. Tell me more about both the data creation side and also the synthesizing that data into turning it into something that can help you make a decision to respond more effectively.

[3:00] Rob Vernon: Yeah, so our background as an IoT-style thinking company has always been about aggregating disparate sources of information together to make some sort of informed management decisions. So for us, our work in forestry had been the deployment of soil moisture and fuel moisture and hyper-local weather to feed into models for prescribed burning. That's in essence where we got started. From a synthesis point of view, you've got to have that quality ground truth information for the models to be worth a pinch of anything at the end of the day. That's no different when you think about Earth observation from space. The models that are good are great until the next one comes along. We definitely see our role in that as providing that in situ quality information for improving models based on what's actually going on in the environment.

[3:47] Kevin Sofen: And then you touched on this, but the idea of, again, there's these different types of data then being created. The data is being synthesized in a certain manner, but then talk me through kind of an actual response on the ground, whether you are a fire department sitting somewhere in sort of bush wildland fire area that's at risk, like is this helping you know exactly where the fire is this helping you understand where the potential risks might be happening so you can pre-deploy assets? Talk to me through some of the more practical benefits that this provides to first responders on the ground.

[4:19] Rob Vernon: Yeah, so all of the above in terms of what you said. So detection for us is multi-vector. So we're using sensors to smell for smoke. We're using cameras to visually look for smoke. The Earth observation stuff, the satellite stuff, is another version of that. So where we're going with this is, okay, great, we get a detection from one of those vectors. If we have the opportunity to validate that from another, I, you know, if we see it terrestrially, can we pinpoint it from space, you know, can we do triangulation? Those sorts of things. So in the essence is about getting a quality detection based on a number of factors from that, you know, the service that we're offering, we're in essence, providing a level of triage around that, you know, pulling the data sources together, sometimes manually, sometimes automatically based on what's available in that given situation, but then we're passing that information on now that works in different ways, depending on, I guess, the fire management maturity of the organization we're working on. So some of our forestry companies, they have their own fire management team, right? So in essence, our triage passed on to them as that escalation point. The other point for other customers is, yeah, great. We want to know about, but can you just call triple zero for us? So triple zero being the emergency services here in Australia. So the management response is somewhat tailored to who the first line customer is. I mean, fire impacts more than just someone who owns the patch of dirt on which the fire is. There is a much broader response and communication piece around that. We see ourselves as very much a part of that.

[5:46] Kevin Sofen: That's awesome. And then the path forward with XPRIZE, I mean, I know there's testing and kind of deploying sometime in the next six to 12 months. And then, so I'm curious to hear your thoughts on that and then also what after XPRIZE? So tell me XPRIZE and then what after XPRIZE?

[6:01] Rob Vernon: So I can speak to our own timeframe in that our next milestone from XPRIZE is in March, in terms of into the next phase of the competition. But our intentions are that by Christmas, we've field tested our, we've already got operational services detecting fires. Some of our teammates have already got operational services using drones to start fires. So our existing customer base affords us the ability to sit with them and go, let's do some cool stuff. So our intention is by Christmas that we will be doing terrestrial detections combined with the automated mission planning around the drone deployments purely just for situational awareness so we're going to detect a fire and then just go and stick a drone over the top of it you know combination of optical and thermal cameras start mapping the boundaries of the fire as it exists at that time so that's that's a service that we're looking to have available to our customers by Christmas. Competition aside now the competition in terms of XPRIZE where we're going next is the intent for the competition is to go and put that fire out now there's some super serious physics that goes into that. And we're up for learning more about how we might tackle that. There's a, so track B, the competition is what we're in. we're the only team left that's been led from the Southern hemisphere. So we've got to meet up in October in Salt Lake City. We're going to meet some of the teams that were left in the competition and see where, see where we might be able to cross paths in terms of, you know, actually solving for putting the fire out. But as I said earlier on, the intent behind all of this was we're all businesses, most of us are startups where we're interested in the business of doing this and the public good that goes along with it as well. So yeah, there's a long and exciting future ahead for all involved.

[7:40] Kevin Sofen: I mean, it's a cliche phrase, but it does take a village to address this problem. And I know that even with an XPRIZE ecosystem, there's Track A, Track B. From a high level perspective, what is Track A? What is Track B? What's the differences?

[7:51] Rob Vernon: So Track A was the detection from space piece and communication under 10 minutes. We've got that capability, which is great. And, you know, this conference has been great to be talking to a few others who have literally got birds in the air or plan to have additional birds in the air. So that's fantastic. Track B is the automated response. So the go and gump of the out piece. So that's, for us, it's quite a holistic piece from Indicium's perspective. We are the orchestration engine. That's going to be responsible for our customers in addition to solving for terrestrial detections. solving for the Earth observation detections and then automated response. So it's an exciting place to be.

[8:28] Kevin Sofen: Yeah, so it's Indicium.

[8:30] Rob Vernon: Indicium. Latin for information.

[8:33] Kevin Sofen: Okay, I think my third time I got it.

[8:36] Rob Vernon: You got it. Everyone gets it the third time.

[8:38] Kevin Sofen: Yeah, third time. Or else I don't get my own Tim Tams here. The part of the pod is everyone gets Tim Tam. And we got some, I hear it's double coat or bust, so. Excellent.

[8:48] Rob Vernon: Got the double coat. I've done about 10Ks just walking around this conference, so I'm probably due for it.

[8:53] Kevin Sofen: You're due for it, Tim. Yeah, for sure. So kind of final thing is any sort of quote that you live by or kind of maybe even challenge or anything that you're kind of thinking about that whether it's in personal life or in work that you could leave us with here today.

[9:06] Rob Vernon: Yeah, so there's a fundamental thing that I quote often at Indicium and we're a sensing company, right? We're agnostic to devices, we're agnostic to comms. So our customer base, there's this fundamental theme I guess climate and climate change in everything we do. Fire is just one representation of that. So in essence, you know, the sort of catchphrase for us is measure what matters. Yeah, in whatever context that you're interested in measuring something, in essence measuring it in order to baseline it and then probably change it. But you've got to figure out what that is that you're actually, what matters to you as an organization. So for us, you know, that measure what matters is sort of the key catchphrase.

[9:43] Kevin Sofen: Well said. Well, Rob, you're doing good work. I'm really excited to see what happens next with XPRIZE and Indicium. Got it. They do it. Indicium. Rob, again, thanks for your time and I look forward to staying in touch.

[9:53] Rob Vernon: Mate, thank you. Appreciate the opportunity.

[9:55] Kevin Sofen: Appreciate it. Thanks for listening to this episode. alongside the Institute of Fire Engineers during the AFAC show. If you found value in the conversation or learned something new, we'd love to hear from you. Reach out, share your thoughts, and let's keep the discussion going on how we can shape the future of smart firefighting.