When Disaster Strikes: Drones to the Rescue | Skydio Ascend '24 Breakout Session
Ryan is the Unmanned Aircraft Systems (UAS) Coordinator for the Alaska Department of Transportation & Public Facilities. He leads efforts to safely integrate and advance UAS and remote sensing technologies across Alaska. With over a decade of experience, Ryan has been instrumental in streamlining aerial data collection and rapid response solutions for both public and private sectors since 2014. His innovative approach has driven significant advancements in software automation, long-range communications, weather-resistant technology, and the development of essential policies for beyond visual line of sight (BVLOS) operations and counter-UAS measures. As a Certified Mapping Scientist specializing in UAS, Ryan's extensive expertise includes unmanned photogrammetry, LiDAR, GIS, emergency response, and critical infrastructure inspections.
As extreme weather events become more frequent, the utility and transportation sectors must adapt to ensure resilient and efficient disaster response. In this session attendees will learn how remote monitoring and automated technologies can improve response times, enhance situational awareness, and improve worker safety during storms. Case studies and real-world examples will demonstrate the significant benefits of remote operations in maintaining service continuity, minimizing damage, and developing pre and post-event assessment strategies.
Afternoon, everyone. My name is Mike Brubaker. I lead our commercial team here at Skydio.
Most of us wake up every day most of us wake up every day with a plan for work. Some of us don't.
Unfortunately, there are a lot of, industries that get just ambushed with, bad things that happen every single day. And these industries are, up here on the stage with me, all throughout the audience.
They typically are the the industries that keep our civilization running. Right? They deliver power to our homes, they keep the water running, they, keep their communities operating efficiently, and they are just inundated with different types of incidents and events. So So today, we're gonna be talking about how these organizations, specifically, a utility organization and a, a Department of Transportation, use UAV capabilities in our solutions to respond to these incidents.
And I'm thrilled to have a couple of industry professionals with me on my left hand side here that are going to talk specifically about how each of their organizations do that, and let me let them introduce themselves.
Yeah. Thank you. My name is, Victor Migliore. I'm the operations lead for ComEd.
We're one of the utility companies under the, Exxon brand, and, you know, we're using drones every day. Right? We have our storm pilots that are on call anytime we have a storm, and they're ready to fly.
Thanks for having me here today too. Ryan Marlowe, Alaska DOT.
We, we oversee about a hundred and fifty operators and aircraft across the state. And when you think of the state of Alaska, I always just tell people, imagine the entire US. We have people spread from Florida to California, all the way up to the border.
That's a lot of land to cover with a very few amount of people. So for us, UAS has been an absolute game changer on the ability to manage such a massive state that's constantly changing with climate change. So very excited to showcase that here today.
Awesome. Thank you.
This comment here from one of the FEMA FEMA deputy administrators is pretty timely, with everything going on back east right now. And, it talks a little bit about the proactive approach. Right? We just come out of the gate and expect for us to respond to these incidents without having a plan. So Ryan and Victor are going to a little bit of how they are planning today and have been planning for their next incident tomorrow.
It is no secret. Storms are increasing in frequency and in magnitude.
Just look at what's going on right now. And in fact, over the last, three and a half decades or so, they've increased at a, rate of about thirteen percent.
We prepared this slide several weeks ago when Hurricane Beryl was still on our minds as the topic, and that is just not the case today.
It is now hurricane Helene, which is getting ready to slam into the southeast. In fact, many individuals at this conference are sort of packing their bags early and getting ready to head back to, storm response mode, incident response mode to get back east prior to that storm and get boots on the ground. So that is real world impact today with people at this conference.
And yes, these storms and other things cause massive amount of damage both on the economic side and on the human side. It's not just storms. Right? We talk about incident response. It'd be, man made incidents as well.
It could be train derailments, fires, then there are the other types of things, floods and, hurricanes and the, you know, the wildfires.
Sometimes, you know, somebody has an accident crashing from pole. Right? Stuff like that.
So yeah. Everything from that just world climbed into transformers to the next hurricane. So, Victor, I'm gonna start with you and hand it off to you, and I'll let you take us through your side.
Yeah. No. Of course. Thank you. Yeah. So, as as Mike said, you know, we are constantly having greater storms, and they are affecting our utility industry, more and more frequently.
Right? One example is, back in, July, right, we had over twenty tornadoes of our service territory in Northern Illinois. And how do we respond to that? Right?
And, using drones has been incredibly helpful.
We have, again, several service territories.
I can speak for ComEd a little bit more than than the other, utility companies, but, it's a lot of area that we have to cover, and I'm sure, a lot of the utility companies face that same issue where it's like, how can we respond to a storm event faster, get people out there faster, but also let our customers know that, hey. We know what's going on.
We're working to resolve the issue.
We're trying to get power back to you as quickly and as safely as possible.
Did you wanna take us through in the audience through, who who Exelon is?
Yeah. Yeah. So, Exelon is, you know, again, one of the largest utility companies in the US.
We have several different, Opcos. My background is with Fed. We have, Go, a BGE, Pepco in Atlantic City. So think of the northwest corridor going DC to New York.
Do you wanna drive?
Yeah. Yeah. Thank you. Okay.
So, we've had, you know, more frequent and more intense storms.
And, with that, it's been, very, very difficult for us to be effective in resolving a lot of these storm events and a lot of storm issues. So one of the challenges that that we've had is how do we go out and we respond to these issues in a in a quick in a safe safe manner.
Right? And as you can see, right, with those graders, we have more dam more incidents, crews that are out trying to down line, down poles, and remotes. And it's it's quite a challenge. And drones have really they've opened up a whole new avenue to identifying a lot of these faults.
We have obviously, we've worked on establishing our our drone procedures, and we have worked on pretty much setting up a drone emergency response, playbook to help us identify how do we go out and how do we dispatch our, crews, how do we dispatch our pilots, how do we, make sure that we use drones as effectively as possible. Right? Because there there are some scenarios where using drones are better than others. Right? So, you know, a team might be able to identify an issue using a thermography camera that's on the ground versus using a drone. So it's it's all about establishing a good process and procedure in order to have our storm response teams have that have drones go to the locations that we need to, when they need to be there.
So, the way that we try to establish that is with what I call a pre preflight checklist.
So, this is BG and Es, but we use Microsoft forms as well. But we try to identify, okay, how are we going to use these drones, where are we going to send them to, and how, you know, how is it going to be utilized in the most effective way possible. So, as you can see there, right, we primarily focus on locations that are difficult to get to that might be a safety hazard for our employees, and that might be, you know, something that, okay, the drone isn't only gonna save time, but it's also gonna be safer.
The big four, damage assessment, things that we try to tackle, right, are poles down, wires down, vegetation, and to reach locations.
So, poles down. Right?
Again, if it's in a location that's that's difficult to get to, super easy to just pop up the drone, fly down the feeder, locate where where that is, take a picture of it. Or if you have a livestream, right, be showing that livestream feed of of what we're seeing. And, of course, my background is in aviation, so it's usually a lot better to have a expert that's there real time with you, like, saying, okay. This is, you know, this is what supplies we need to send out there. This is the crew we need to send out there.
This is, you know, how long it's gonna take them to get there and and get an estimate on the turnaround time for resolving, the trouble.
Right? When we were, talking ahead of this presentation, you were, expressing some of the challenges you have with acquiring, more skill set, you know, skills based workforce. And, you're thinking about the potential use of remote operations, a big theme of this conference today. How are you guys thinking about how that remote based operation can impact the big Ford assessment?
Yes. So, again, right, the skills based work for it. Right? So, a lot of us were having issues finding, you know, people that have, you know, amount of experience out there and do the repairs and identify the repairs.
Right? And how do we resolve that? Right? We still wanna keep our our workforce.
We want to make them more effective. I think one of the biggest concerns that that people have a lot of times is, oh, drones are gonna, you know, replace, our jobs and and things like that. But, no, it's a it's a lot about trying to make, the current workforce that we have more effective. And so, we combine not only dock operations, right, to go out and try to identify faults, but then we also try to use artificial intelligence and machine learning platforms to then go out and and build, you know, confidence and build a a system that we can identify faults without having any sort of human intervention. So, it is something that we're working towards, and it is a long term goal.
And I know, you know, today, we're we're making some big progress in that, so it's, it's very exciting.
Some of the other use cases that we have. Right? Social media.
Again, as I mentioned earlier, are we have people call. Right? They're asking, okay. My power is down.
Right? When is it gonna be back up? Right? I have, you know, I I'm out in the field, doing storm response, storm patrols.
Right? And people are like, hey. Can you use my power? Right?
And and so and and doesn't necessarily work that way. Right? But it how can we more effectively communicate with our customers that, you know, we're working on it. We're we have our crews out there.
We're being not only efficient, but we're being safe. And and part of that is through social media. And, drone imagery goes a long way in doing in in doing that because you can now see, oh, okay. This isn't just a branch that fell, you know, on a line, or this isn't just one pole that's down.
Right? It's four poles. And so, we we think that's that's a way more effectively communicating with our, customers, but we're also, you know, for damage assessments. It is very, very, very useful.
Again, as you can see in that image there, this is somebody's backyard. Right? And, to go there right before, you have to potentially walk in an environment that you're not familiar with. Right?
You're potentially going to somebody's property that may not want you there. Right? And so, it's a lot easier to just, you know, go to somebody's front door, knock, say, hey, you know, we're gonna we're gonna fly a drone here, try to identify, you know, why you're out of power. And it just it makes things significantly safer. And safety is, you know, the biggest aspect of why we use drones.
When when I when I got into this industry several years ago, there was, you know, a response from the consumers and the customers about the use of drones. How has the demeanor changed over the years?
Yeah. It's it's improved a lot. Right? And so I remember, myself as well. Right?
Ten years ago, and and we would have, these these open discussions with the communities about, going out and utilizing drones for for different use cases. And people are like, oh, you know, first drone I see, I'm gonna try to shoot it down. Right? And, thankfully, that's that's improved a lot.
Right? And people, you know, if they see us in uniform or they see us with our high vis vests, hard hat on, right, saying, hey. You know, we're gonna go out there, and, you know, we're we're trying to actually help. We're not trying to, look in your house or things like that.
And so, and I think maybe it's not just a a a time thing, but I think people are you seeing the different potential use cases of, drones. And so when they see that, okay. This is actually trying to help my situation, not not make things worse for me. I think that helps out a lot as well.
Yeah. That's great.
Again, we also use, our drones for post storm feeder patrols. So, we have specifically at ComEd, about fifty percent of our pilots are part of the reliability, inspection team. So they will go out there, and they will do their storm, patrols, making sure that, okay, there's no, you know, high priority faults that need to be taken care of. And then, during the storm, we'll also do, staging.
Right? And we'll either do mapping or what at what we do at ComEd a lot of the times is a live stream. So, you know, they're getting ready to stage trucks at, you know, a shopping mall or at a garage. And so how do we share that information with our, OCC or our our command center.
Right? And we've noticed that a way of really effectively doing that, right, is just, okay. I'll go out there, launch the drone, start a live stream, and there it's there. Right?
There you know, if somebody has a question, I'm on the field. I'm I'm out in the field. They can just talk to me as if I were there with them, and they can see in real time what I'm seeing and provide any feedback on, you know, like, hey. You know, can you, fly over this area, or we need more room for more vehicles, things like that.
Yeah. I think it's, maybe hard to comprehend the kind of impact you can have when you're staging literally, you know, dozens or hundreds of trucks, for example, coming into a response territory. Can you, like, just try to relay the impact of that upon your organization and the public Yeah. And how drones could help that?
Yeah. Yeah. Exactly. And and it's it's, quite a lot of vehicles and staging them, it it it's quite a it it's not something that you just do right then and there.
Right? We plan for this stuff months or even years ahead of time. Right? The best staging locations for if there is an incident over in this area or if there's an incident, you know, in our northern region at ComEd or if there's an incident in the West region.
And so we try to work with, the communities as well for staging the, the vehicles during a storm event. However, right, maybe there's some information that we didn't know or that we weren't familiar with. And so having a drone pilot that can go out there and broadcast this information is extremely useful.
So we still use, our traditional damage assessment methods, as well, alongside with using drones. Right? It's not something that's one hundred percent one or one hundred percent the other. We do leave that for our, UAS pilots to, try to decide on their own.
Right? If maybe it's better to just take the drone up, take a quick picture, come back down, or maybe, you know, on a lot of our damage assessors vehicles, they'll have a PTZ camera that they can just point. You can see there, it looks like that insulators falling off of the cross arm. Right?
Okay. That might be something that I don't necessarily need a drone for. Right? But then we also are trying to integrate automation more in trying to use that alongside our traditional methods for identifying faults.
So, again, we're kind of stage one to two. Right? We started doing one to one operations.
We're hoping, you know, now with the X10 doc to get to one to many to where we have one operator in a command center operating multiple drones at the same time. And ultimately, we do wanna have, you know, the goal of having a drone go out anytime we have a storm. Right? Identify based off the data that we have, what feeders we need to go to, and go ahead and just automatically do a scan without needing a pilot there at all. Right?
Could you maybe, talk a little bit about how your organization is looking to then take that data, process it in AI, and specifically use AI for, anomaly detections for damage assessment?
Yeah. So, it's, definitely something that we're working towards.
We are using, software right now to, build and train to identify these faults. Right? I think there's a couple of things that we're really focused on. So cross arms, conductors, the top tops of poles, and we're trying to train our machine learning, software on that data, feeding it more images so we can build, you know, the confidence to where okay.
You know?
Some cases, yeah, you'll still need to have a human look at that image and say, okay. Yeah. That's a broken cross arm. But a lot of that work can be identified by machine learning software, or, you know, it might even identify something before we can even get somebody out there. And that's that's really where it helps with a lot of the preventative maintenance and a lot of, making sure that we're resilient as a utility company, making sure that, you know, we have the best preparation possible before there is even a storm.
That's the the get back up fast that Adam was talking about earlier.
Yep. Exactly. So as you can see here for, BG and E, right, we've identified, our substations and where in the future we would like to set up our drone docks. And you can see it covers eighty percent of their, service territory. So it's something that with the Beyond Visual Line of Sight Waiver and with the drone docs, we could not just only storm response. Right? Tons of different things.
One that we focus a lot at ComEd is substation inspections, which, again, we talked about earlier, this morning. But it's it's hopefully, it's a game changer in that aspect. And we hope that not only with the new technology, right, but with the changing regulatory environment, we can then go out and we can have these docs improve our safety and improve our reliability.
So with BG and E, they've talked about, the eighty percent number. That figure changes a little bit as you go into your service territory specifically. Yes. Not quite as high just due to longer feeders. So as the geography changes and the assets change, that gets impacted.
Yes. That is correct. So, you know, again, right, it's a new technology. Right? But also the regulatory environment, we might be limited in in the distance that we can fly. And so, again, yeah, the distance between the substations might be greater than the distance that we're allowed to fly or that the drone is capable to. And so that's why we're closer to approximately sixty percent for, if we had a drone dock installed.
How do we get you to eighty percent? Like, what's the magic number for distance?
You know yeah. Exactly. I mean, you know, we you know, maybe fixed wing. Who knows?
Right? We we we gotta test it out. Dock, x one hundred. Right? Yeah. So we'll see.
Next year.
Yeah. Yeah. Next year. Next year.
Next is that. Yeah.
And then as well, so we we talked a little bit about, how the traditional damage assessment works, But with the drone in a box damage assessment, right, it's the same steps but now we're using the drone in a box, right? And so we're going out there but fully autonomously trying to identify, the fault, assess it, and then have the machine learning software then give a priority to that fault, like, when does it need to be repaired, and then send out our truck rolls and send out our crews.
Awesome.
Alright.
Victor, thank you so much. That's super helpful. We look forward to to, rolling those out and and getting to a better state for you guys here. Of course. Right?
Thank you. Where do I start?
Natural disasters. Yeah. You saw during today's keynote, if you guys watching it, where we're deploying drones at, all across Alaska.
We have had our fair share of disasters. I think since our UAS program, we've responded to sixteen large scale disasters. That's earthquake, erosions, avalanche coming down through neighborhoods, landslides here lately. And so, really, a lot of this technology, you know, we're taking multiple pieces of technology out there, but they're finally all complementing each other.
You know? We're we're out there with Starlink, and we're able to livestream from places that we've never had, you know, let alone a two meg connection, but broadband connectivity. So, what we're looking at is is the ability to scale remote operations, the ability to share that load. I talked earlier about we don't have enough staff to be everywhere at once, but we do have enough technology.
And so if that technology can be pre pushed out there, the ability for us to log in and remotely remotely scan, remotely operate, share that load to collect that data, that is the key item that we've been working towards, and it's it's being successful. The x ten platform as part of the the USDOT smart grant, which ours is the aero project, we found tremendous growth and and are ready to scale that model going forward. So, we call it our one seventy two rule, which is basically whatever you can fit into the back of a Cessna one seventy two is what's going with you. And so now that systems are getting smaller, we can bring a survey kit, an x ten, a Starlink panel, and one person can do the job that it used to take a team of ten.
So that is really changing. And and as we look at docks too, reducing our carbon footprint, a lot of our operations, if we're going out to the Aleutians to go do mapping missions, you know, that's that's a sixteen to twenty thousand dollar flight to charter that jet, get the team there. So we're really excited to see where scaling and and having these systems on standby really make that difference.
So Yeah.
And I was fascinated to hear, like, that you're putting these into the hands of your communities in these remote locations, and they're doing the flying, or they're powering it on with remote flight deck and you guys are flying. Correct?
Yeah. So we've been working with, tribal a lot of tribal governments. So Skydio has been coming out and will train the community. We've been partnering with Esri to do a lot of the back end processing, but really just showing is you don't have to have the knowledge there.
It's shared. You know? Everything is going to the cloud. We're processing. We're really allowing for this to kind of ebb and flow depending on the needs.
So, that has made a big difference as far as that scalability is. It doesn't not everybody needs to have a fancy you know, processing laptop when it can just be done in the cloud. You can continue on, jump back on your snow machine or your four wheeler or boat and head back to the office, and that data is finished when you get there. So if that is, you know, something that didn't exist a year ago is now normal routine for a majority of our operations.
Avalanche, dock based monitoring.
I'm sure you guys have seen, triggering, but this is really where we're heading is the ability for, events triggered and remote monitoring.
The ability for dock based operations, terrain mapping, this really for us is the game changer of allowing for mitigation, allowing for us to see just how much snow volume we have at a lot of these places. With a lot of the inbound AI and snow depth modeling, we can even see a run out model of, hey, there's a ninety percent chance that we have enough volume that'll hit the roadway. So now it's alerting the public earlier saying, hey, Heads up. They're, you know, avalanche.
Don't stop in these avalanche zones, or we'll go ahead and mitigate pre storm so that way we know that the potential isn't there. So all of these things, those analytics of having these these these systems available are really changing. And I I've been working on this for a while, but think of it as like an office printer. You know, everybody has access to it in the office.
You send it the craziest things to print, and it just takes care of it. That is what we're seeing with docs is these missions within the bowl, or that, you know, that that defined area are allowing for us to capture that low hanging fruit data, not just whenever you need it, but daily, twice daily. That type of change detection as far as avalanche mitigation monitoring. But not only that, but did somebody get blown into the water?
You know, how do we respond to that? We've been seeing a lot of groups work towards that drone as a first responder and seeing the the live demo today. That is exactly what these four are are for is how do we respond to make sure everybody is okay, but documenting it moving forward.
Do do you all feel that you've been able to avoid a tragedy as a result of with this particular use case by getting out in front of it?
It's changing well, what we've learned with Avalanche has directly applied to our ability to respond to landslides.
Same same material, just different seasons. So, that everything that we've been doing building up to this ability to operate in these harsh climates has allowed for us to operate in torrential downpour. But being able to say with certain, you know, fact, that's that slope is stable, and we've never set foot on it. That's an an incredible, you know, use of this technology that you're you're you're you're passing that onto the public and saying, hey. These areas are okay. We need to watch out for these areas. So that's really where a lot of these tools are just doing the purpose is keeping folks out of harm's way.
And in the past, you would have sent somebody out there climbing.
Survey yeah. We'd have a survey team of twenty out there figuring out if it's, you know, stable, is in the rain. It's like communication based. So and most importantly, remotely. We've a lot of these sites, we've been logging into over two thousand miles away and being able to operate these systems daily and capture that data, and it's automatically uploaded, processed, and ready for us in hours.
And the financial impact of that is?
We we rough numbers right now, but I talked about that earlier. A couple of these sites, when you think of a plane ticket for most folks, you know, it's a couple hundred dollars. For us, most places, it's eight hundred dollars, even out of the Aleutians, couple thousand. But if we're traveling with hazmat batteries, now we're chartering a private jet, so that starts to jack up pretty high. That carbon footprint as well, that's really what we're seeing is the ability for us to have these systems is reducing a massive carbon footprint for us as far as operations because we have the technology and the remote sensing tech installed. So cool.
Speaking of technology, I love to make diagrams. So if you've ever worked with me, I love to share just how does this work. You know, you see the dock on stage, but is it turnkey? Yeah. Sort sort of say.
But this is what we've created, not only on the security side, but the ability to scale this. You know, the dock requires a weather station.
The dock requires security. We've used ring cameras. We've used all these different items. You know, how do you see I I think I get alerts.
You know, actually, I got an alert last night. There was a deer at one of our sites, you know, a human detected, and so sure enough. But all of this stuff tied together is what makes that operational, response capability there. We we've been focusing really, you know, really a lot of this technology has been dependent on low earth orbit technology.
So the ability to now have broadband communications anywhere. So we've been repurposing a lot of signaling signal cabinets and using Starlink.
We've been testing out some of the VHF rebroadcast. But how do you install everything that you need to operate there without being there? So that's been a really cool item is if we're operating at some of these landslides, we can talk to air traffic, we can broadcast that we're operating, we can repeat that, the ability to watch for ADSB. So all of these things that we're learning by installing these across the state. The other big one is security, IoT security. How do you operate that system remotely across space laser?
How do you keep that that item encrypted? How do you do that? And so that's been a big focus on our point as as we start to scale these technologies, you have to think of it as like a good old LAN party days is how are we gonna connect all these as fast as possible, have prioritize video streams, and how do we process that data, and get it to the cloud and get it to the end user in a record time.
So You've done a ton of yeoman's work on, you know, building all of this out. Like, if if somebody in the audience here is taking a step towards this for the first time, what's the one thing that you're gonna direct them to right now, first and foremost?
Reliable power. We'll start with that. If you don't have power or you don't have the ability to make power, I know the California folks, you have beautiful sunshine. We're not as privy to that. So, solar. I know, you know, Caltrans is working on, off grid docks and solar powered docks. I think Velocity is here as well with some of their docks.
So Nice plug for the vendor outside. You can take a look at the box outside.
So really, it's just like that's really what you're looking for. Second item, communications. For us Yeah. You know, we've never had access to fiber, or a lot of these things off grid. So to have this technology, not just one, but we've been starting to experiment with bond, of these. Just like you can bond LTE, you can bond Starlink panels for for higher throughput. So really, overall, we're just recycling all of these technologies that are available and and making them useful in the field.
So Thank you.
And then, of course, the data products. This is my favorite part with a lot of this is how do we tell that story? How do you know, that investment of that dock that we've now installed in the field, what does that give you? On the left here, the the snow distribution maps for avalanche data.
That is a key item so that we can either do, drone mitigation, where we're dropping explosives, heli, or even howitzer mitigation, all of which, that's a pretty precision map of we should target the upper, cornice so that it drops down and hits everything else. The middle picture, we're talking about landslide. We showed at a place where we had zero connectivity. We didn't have any GIS data.
We didn't know what houses were affected. We're building all of these things live in the field because we're saying, hey. Here's the area. We need to know roadway information, all structure information.
We need lots. We need to know who's what. That's where that teaming in this live situationals. We're putting that imagery into the flight deck, putting that imagery under the control.
Everybody has that information time. That's where a list is coming together. And then back production. The ability for daily as builds.
You wake up in the morning and the project scanned. Why did that contractor charges for sidewalk aren't installed yet? That's a visual representation of something that normally wouldn't catch until we'd already build for it. So all of these things that are based off of Imagity and all these analytics can now be done with these remote sites.
And then data workflows. This is a big item. Once again, another one of my charts, but this is how automated data works. We've been focusing really hard on when I press go on that dock, what does that mean?
That means that I expect that dock to not only operate remotely, but operate that mission plan to capture the imagery at the ground sampling distance I've preset to then dock and start uploading, call out to site scan, where we use Esri quite a bit, and then basically transfer that information up to site scan and start that processing. Now for multiple sites, which is really where dock pays off, if we're flying every day or morning and PM, now we can start seeing rapid change detection. So if we're able to see, hey, the river's moving at this speed, we're losing twenty feet of of shoreline, you know, at this, perfect.
Order eight more trucks to start bringing in material. That type of information, and especially with this, the Dalton Highway, our teams at the SEOC had better information than the contractors in the field, and we were sending vehicles. They're like, we didn't order these. I'm like, well, yes, but we could see that you needed them.
That's a crazy change of pace where we had just got Starlink, like, couple days before this happened, and so we're like, stick it out there and see what happens. And it just happened to have everything that we needed. Even on the live streaming, where we're we're live streaming, we're talking about the events, and somebody pipes up and they're like, where'd you get this video from? And we're like, well, we're live streaming it live in the field, and you could just hear the pause in their voice of, oh, I guess that's normal now.
So, you know, some place that has been offline forever is now accessible, and that information can be shared immediately.
And documentation, even with window surveys with FEMA, we're not collecting data and sending it to them. They are in, you know, on the Teams meetings. They're in the field.
They're watching this live, and we're all making these decisions today.
Which is a huge business problem in and of itself to pull all of that together, and you're I think you're gonna talk a little bit about that.
Absolutely. So and when we're looking at this data, the big item for us is how do we collect this data as a structured environment as as a team and have it all flow? So as we're collecting the imagery, as we're conducting, we're showing up, we're following under incident command system. All of that is being even all the way down to the resource request from the community, we already have a lot of these prebuilt so that we're just filling in the information and submitting that for FEMA for reimbursement.
So that's a key item. You talked a lot about preplanning, you know, prestaging assets, prestaging technologies, pre filling out all these forms so that when I'm long gone, somebody else can just say, oh, populate it in. So and same with docs too. We currently have it as if we know there's a landslide or some disaster, it's getting a dock, it's getting a Starlink panel, it's getting a radio.
It's already prewritten. We know the cost for it, and that is getting dispatched immediately. So those are the key items of once you build that tie in and once you build that workflow, super crucial to tie those things together.
And prior to a drone program, how did you do that?
You miss filled out your paperwork, and you didn't get reimbursed.
That's what happened. So when we're looking at cost effectiveness, we're fixing a lot of air, the human air factor in that. And it's like, there are a lot of states that, I mean, you have to eat those costs. So we're fixing that gap with data that we've never been able to capture before. Yeah.
That's awesome. I don't know.
Yeah. Yeah.
I have one final question for the both of you, and I think we'll open it up to questions from the audience.
Three years out, five years out, what is, like, one problem that you were really hopeful that you can solve with remote based operations? Victor, why don't you take that first?
Yeah. No. So, definitely, being able to identify faults and and identify, hazards as as quickly and as efficiently as possible. Right?
So the the one program that I dedicate a lot of my time to, not related hundred percent to to storm response, but is a substation inspections using the dock. Right? And so, our goal is to have one dock per substation in the future. But not only will that help with just the substation inspections, but just a a lot of data collection and data that we can use for things that maybe we're not even a hundred percent sure that we needed data for before.
Right?
And so, things from, you know, identifying a fence line security that we talked earlier to, again, looking at snowplow routes or looking for, looking for a construction in the substation, construction nearby to where our property is, things like that.
So Oh, I gotta choose my favorite ones.
So Just one? Yeah. One.
Installation of docks on our critical assets. And what I mean by that is, you know, Alaska's split into aviation, highways, and marine ferry terminals. So docks that sit on airports that can collect, you know, and and and document that schema and say, hey. Something's out of out of line or docks that sit on our ferry, and inspect that ferry either in transit or inspect the port of call.
Really looking at these things holistically, but not just DOT having access, having the entire state and and multiple agencies have access. So that's really where we see this back to my office printer concept. Anybody in Texas should be able to print, you know, and use these docs in Alaska. That that back and forth and from what I've seen, you know, even today is like, it's here.
It's like we're we're getting ready to see the shared use of we can have assistance from anybody in the world and have that information immediately.
I feel like other state governments should look to Alaska as a North Star for some of the technology you guys are deploying. The incredible, use of your drone program, Ryan. Thank you so much. I'm gonna open up to questions from the audience, and Tommy will pass the microphone around for us.
Yeah. Hi. Thanks for that wonderful presentations.
I have two questions.
First, I I wanna get a feeling from you about what are what kind of sensing devices are being developed? We see a lot of visual images because we're very visual animals, But what about other kinds of sensing beyond the electromagnetic spectrum that we see, but other perception devices that would be good for evaluating electrical faults or or avalanches or other problems. That that's my first question. And my my second question, has to do with actual intervention.
I mean, how much are you using drones? I heard you mentioned using explosive to set off avalanches, but clearly can you use drones to intervene so you identify a potential fault? Can you use drones to go in to remote places that are really hard for people to get to and use the drone to actually cut a tree limb or or fix a device?
So I'd like to hear you opine on both of those.
I guess I'll take the first shot at it.
First item is IoT sensors. You know, I I view drone as another IoT device. It's collecting data, we're looking at it, but we're also seeing, sensor based technologies, hydrophones.
In Alaska, we have our ShakeAlert system. We have a lot of these systems that are out there, in many cases, legacy systems that when you're saying, hey. What triggers this event? We're taking those items and say, hey. If this happens or an earthquake is detected here, pop the dock up and go take a look at stuff. Start collecting that data.
On the Avalanche side, we've started installing hydrophones, that vibrate when that much snow mass is moving, which is an early indication that the Avalanche is happening real time. That then would trigger the dock operations and allow for us to respond or at least alert the public, hey, an avalanche has occurred at this location. Heads up, secondary, and third ones are inbound. So that's really where we see IoT devices.
I don't know too much on the Skydio side as far as what's being integrated more onboard the platform, but a lot of this environment around on the outside is what's adding to those event triggers, what's a company that weather based. You know, we're we're starting to see virtual, VWOS stations get installed. Instead of having these only sites at the airport, people are starting to dispatch them where we have these microclimates. So, really, the Internet of things and tying all these things together are allowing us to have that better picture of what's actually happening out there, which is really all we're doing is just trying to figure out what's happening so that we can make those informed decisions.
So, what was your second question?
The sensors.
Sensors. What do you do to intervene?
Could you fly in during a storm and cut a branch that you can see as wack Got it.
I'll pass that back over to you.
Yeah. Physical repair system when delivering things that people can't easily get to be able to do these things?
Yes. That that's a great question. I'll I'll answer that that part, and then I'll go into the different sensor packages. So we we're looking at several different things, right?
One of the project one of the projects that Exelon has worked on is installing bird diverters on our transmission lines. So believe it or not, you know, birds, yes, they've been flying, you know, all over the Earth for millions of years, but there's this thing that's come into existence in the last one hundred years that they're not familiar with, and it's a huge environmental issue. Right? Because the bird, they might may or may not see that transmission line, and then that can affect endangered species, but also, you know, having birds on our conductors, that's not good either.
Right? And so, we've worked on installing these bird diverters, and all it is is just a yellow reflector that's the size of, like, a dinner plate. Right? And that's enough for the bird to be like, oh, there's something there.
Right? And they'll fly around our our transmission lines. But we're also working on, you know, pilot programs to potentially use drones for lubrication of our switches. Right?
Or using drones for pulling like a p line to where, okay, we gotta get this conductor across this lake or across this river. Well, you know, before you'd have to either go on a boat or on a barge, but if you can go ahead and send a feeder line first, well, then you can just use that to then tow the conductor right behind it. So, that is, you know, ton tons of different again, like I said, you know, there's probably ideas out there that I know I haven't thought of, maybe somebody else has thought of, and, you know, it's just something that that that we need to work on. And on the, and on the sensor side of things, right, I I like what you said as well about, the part of the electromagnetic spectrum that that we can't see.
Right? And that's how we try to, identify the best sensors for our different programs. So, one of the ones that, you know, I would say, you know, would be really nice. Again, future asks, a lidar.
Right? Our vegetation management team. Lidar is great at penetrating through vegetation, whereas, you know, the three d scan that SCADI offers is great, but that's one of the faults there. And so having a compact drone that has lidar, right, it's it will be a great good to have.
And and, again, you know, with the autonomy that, Skydio drones have built in, it would just make the training side of things a lot more, faster and also more efficient for for our pilots.
Thank you. Lidar hits the live action. Yeah. Great. Thanks.
Any other questions from the audience?
Hey, Ryan. Thanks for, the presentation.
You know, know some of your, diagram showing how you're taking the data and bringing it into the Esri ArcGIS Enterprise and and some more environments. Any, lessons learned or best practices that you found?
As far as lessons learned, be ready be ready to have a you fail a lot. You know, it's just like especially when you're coming into that environment of a stack solution, but Esri has been amazing. You know, we've we've seen some incredible items. I know the video server I'll pitch them.
The video server team is here today too, but that, of course, has brought us into another world of, hey. We have imagery pretty much taken care of. It's processed. It's we're able to access it.
It dumps into a repository. It's available to the public. We can source it back into our maps. Now we're looking at video.
You know, how do you bring video into an enterprise environment? How do you split it out to full motion video? How do you tie those things together? So building on that ecosystem and making it, you know, streaming and and bringing in that that that next data source, I think as far as, you know, our lessons learned is there's sometimes that there's not a solution on the wall, but once you have that idea, it's like that was one of our bigger items as we scaled so fast as an organization that our IT guys absolutely hated us because every drive was redlining, and they're like, you have to figure this out.
So going off network and allowing that to upload directly to the cloud, as I mentioned with Starlink, that has been a major relief saver on a lot of our not only our network speeds, but our ability to get data out and share it as fast as possible. So looking at it on that side, the the shareability and the ease of use, and and even the tie in with other, you know, other processing engines as well, we've seen that across the, Skydio cloud environment of just all of these groups are building tie ins. I know PIX, or GeneX, and and we're we're processing with PIX. All of those things giving us that same dataset, we wanna make sure it's available to the cloud at the end of it.
I, thank you to the entire panel. This question is probably more directed towards Emmy Award winner, Ryan Marlowe.
When it comes to right. You mentioned that, you know, for certain types of events, they get Starlink dock gets deployed. Is there a, like, a quantitative kind of matrix that you use to determine, is it magnitude of damage? Is it time expected for cleanup of this one gets a dock, this one just gets routine normal flights?
Great question. This has been an interesting one on the DOT side is I always try to tell, we are not first responders, but we are because we have this technology, we're coming into this environment to tell a story. A lot of these landslides were showing up to these places as the first one and normally are switching from, alright, let's get video, let's figure out what happened, if we're assisting with victims or any issue there, it's how do you switch from that to then, as you mentioned, long term data collection.
So that's a hard mindset to dump into, but I will say utilizing ICS for those groups that do not utilize incident command system, that is a crucial element of there that you plug into. You show up on scene and basically step in, and you're like, alright. Well, we're here to do this. After this gets taken care of, now data collection.
So that I would say has been the best lesson learned is make sure you have, an ICS system enabled if you are responding as a large scale agency, and it definitely takes care of a lot of those items. Or if you're a single individual out there, it helps with the documentation process to make sure you're just not trying to take care of everything at once because that doesn't do anybody any good. So Thank you for the excellent questions. We're gonna wrap up with a couple from our virtual audience.
So we've had this streamed virtually and had some questions come in. So, one of the questions was for FEMA claims, is there a business case to perform pre disaster critical infrastructure missions so that you have before and after data?
I guess this is me again.
I wish we could say that we could do, and I guess reading that question, like, going out and collecting information before a disaster.
We have do not have that luxury in a lot of the state. I wish I could decide where earthquakes are gonna happen at. But, for the most part, we are responding to events that are near communities where there is loss of life and the items like on that side. So as we're looking at this moving forward, it's really critical to be able to tell that story.
You know, the pre post damage assessment, that window survey, is incredibly important to be able to conduct that correctly and capture that data so that you can showcase, hey. What are we looking at? How how does this change? How does the team move forward?
So hopefully, that answered the question you're looking for.
I I think I can answer at least from one industry perspective, and that is, the utility industry.
It is absolutely critical for them to have records of what infrastructure look like prior to a storm when it comes to recovery costs. So at least from that, industry's angle, yes. The answer is yes.
And if you don't mind going back one slide, I think you can show the actual FEMA policy as of twenty twenty.
Yep.
One more slide.
There we go.
Great. And then I think this question is for Victor. Hamed management been supportive of the ecosystem initiatives. What challenges are you experienced management to buy into large scale adoption?
Yeah. So that that's that's a great question. Right?
Drones, even though they've they've been around for a while, still, a new technology.
Right? And and and to get buy in, it's it's a lot about showing what the future could be, right, and what drones can can solve that we're currently having issues with. And and there's a lot of things that are, again, right, storms that we're having more of them, and they're more intense.
And, again, like I was saying, the skill the workforce that we have, they're great. They're very skilled, but it's becoming more and more overwhelming if we don't grow our workforce, which we aren't. And so with that, we gotta we have to find some other way. Right?
And and automation, that helps, but also showing that, you know, hey. Instead of trying to get somebody that has no background in this, right, and trying to train them versus, okay. We already have the best people that we need. They're already working in in our company.
They just need a a better tool, right, or they need a tool that can help them identify issues, identify faults, help with the storm damage assessments.
That's how we're working on convincing our our management and our and our leadership team to, utilize drones. And so and it is something you know, it's not a it's not an easy road. I I know I see a lot of familiar faces here that work in the utility industry.
We actually we were all at a at a conference a couple days ago talking about the these same challenges, and, you know, it's, you know, I I hate to put things on like a black or white, but it's either the people that are pro drone. Right? They see the the future of it, whereas maybe the people that are anti drone see this as maybe an additional expense that they could avoid. Right?
And it's just trying to convince that one side, like, hey. Yes. This is a greater expense, but would you rather, you know, work on this or, you know, have, you know, customers that are upset because they had to wait days to have, you know, power back home. So yeah.
Okay. I think we are out of time. Victor, Ryan, thank you so much for both being, innovators in your respective fields and, for your time and the presentation. Thank you. Thank you.