Disaster-Ready Tech: How 5G is Transforming Emergency Communications

Experience a compelling journey into the future of emergency communications as James Thomas, CEO of JET Connectivity, reveals how custom 5G solutions are revolutionizing disaster response. Learn how their innovative technology provides critical connectivity during natural disasters, from California wildfires to remote maritime environments, ensuring first responders maintain reliable communications when traditional infrastructure fails.

Discover how JET's robust 5G platforms are transforming emergency response capabilities with their unique low-power, ultra-durable design. James Thomas shares fascinating insights into the challenges of developing technology for extreme environments and explains how their solutions are bridging critical communication gaps in disaster scenarios.

Resources from this episode:

• Learn more about JET Connectivity: https://jet-eng.co.uk/
• Connect with James Thomas: https://www.linkedin.com/in/jamesetthomas/

Listen to the Episode

Watch the Episode

Transcript

James (Host): Hi everyone, this is James from the CTRL+Listen podcast, brought to you by Octopart. Today I am here with James Thomas, CEO and founder of Jet Connectivity. Thank you so much for joining us, James. Great to have you here.

James Thomas: Thanks for having me. I appreciate it.

James (Host): Anytime. Let’s get started with the basics: the company. Do you want to tell us a little bit about the company and its background?

James Thomas: Sure. The company was founded in August 2020. We first started doing any real work in January 2021. It came from an experience I had doing offshore yacht racing. I wasn’t the owner unfortunately, but I was racing in something called the Fastnet. I don’t know if you’ve heard of it, depending on where your audience is based.

We were racing from the south of the UK near the Isle of Wight to southern Ireland and back. We were double-handed in the boat. When we sail off the coast, we turn the engine on to charge the batteries. That’s the normal way you charge batteries, check fuel, and keep things running. When you’re double-handed, you rely a lot on that engine so you can use things like the autopilot, change sails more easily when there are just two of you, and reduce the loads.

One time we went to turn the engine on, and it didn’t start. It was about a Force 7 at that point, so we had fairly punchy sea conditions. I tried to rebuild the engine but didn’t have the parts we needed because something had broken. Things then got worse than we were expecting. The owner had been on deck happily sailing the boat. We had the number four up, which is a small sail, and the number three lashed to the deck. The number three had fallen over the side, and he’d only just realized because it was about two in the morning by this stage.

So we’re two days in, it’s 2 a.m., sailing on the side in a big storm. I’m standing on the bow trying to wrestle a sail back on board thinking, “I’m definitely going to die if I get this wrong,” because we don’t have enough safety systems, we don’t have enough comms, we don’t really know where we are other than by head torch and compass. I thought, “Wouldn’t it be great if my phone just worked and I could tell people where I was if this all went wrong?”

Fast-forward, and I was thinking, “How do we solve the problem of the lack of comms at sea?” This was before the Starlinks of this world really existed as they do now. We realized there was a massive opportunity to build floating offshore connectivity, but it needed fairly custom software. So we started from scratch, built our own 5G stack, and went from there.

James (Host): Wow, that’s a great origin story. Normally it’s, “I was sitting in my garage and drew something up and that became the company,” but I like how it started from such a practical lived experience.

James Thomas: Yeah, I think it’s quite nice to be able to tell people where it all came from. There wasn’t really a single “eureka” moment. I remember initially thinking, “Why don’t we do wristbands on people so when they jump in the sea, they can signal?” and we quickly realized that was a really bad idea. We tested that idea fast and dropped it.

But the idea of doing something for people at sea with communications was solid, so we iterated from there. It was a long way to go from where we started, which was 400-kilo platforms, to our latest ones that are 30-something tons. They’re now huge by comparison.

James (Host): So what exactly is the primary product? What is the main thing produced by your company?

James Thomas: We effectively have two things that we do, which can sound confusing, but I’ll try to make it simple. We build 5G radios. They’re our own radios with our own software, but they’re quite custom, very robust, and low power, and they’re used for really difficult environments.

We’re not using them for something like a “factory of the future.” They’re used for things like disaster-relief connectivity where you might want to run the system off a battery, in the middle of nowhere, without access to normal connectivity solutions.

We also put those radios into our own platforms. We have platforms which are about 22 meters tall—60 to 70 feet—that float offshore. These buoys are there to capture maritime data and provide that data for a multitude of environmental and other use cases.

James (Host): Wow. So what are the challenges that people wouldn’t think of with maritime-based technology?

James Thomas: There are quite a few challenges that people don’t immediately think of. We realized there are challenges in robustness, durability, and making sure the capability actually works long term.

It isn’t just the obvious things like using good gaskets and good seals. It comes down to durability so that things don’t fall apart when they’re constantly being shaken. People talk about harsh land environments, then harsher military environments—and then you put something at sea continuously. It’s salt-sprayed, wet, cold, damp, and moving all the time. It will never be stationary in its life. That makes it a really tough environment to build something reliable for.

James (Host): That makes sense. The movement especially is something people probably don’t take into account. What’s the movement range roughly for one of your pieces of equipment?

James Thomas: The biggest sea state we’d expect to see is about 18 meters. You can imagine what an 18-meter wave looks like. So again, 50–60-foot waves hitting a structure that weighs 30 tons. The shock loads are huge—we’re talking hundreds of tons of shock load from that weight of water hitting it. It’s a really destructive environment.

James (Host): That makes sense. Do you want to break it down to a deeper level of what the individual offering is from the company, in terms of the technology and services?

James Thomas: At the lower level we produce 5G radios. That’s a couple of PCBs we design, some radio cards we design, that host their own software.

I don’t know whether your listeners have much experience with 5G, but we’ve got a core, a CU, a DU, and an RU hosted on the same platform, which is essentially an entire mobile phone network. Normally people just have an RU in the box mounted at the top of a pole. They have a CU and DU on a server rack, and a core that’s on a different server. We’ve put the whole thing on a single board computer. It runs on about 70 watts. It’s a really small-form-factor package.

That’s our core product. That’s what we’re putting into different platforms now. We started with the offshore platforms as the big use case, where we thought the biggest problem was. We’ve since found that a lot of disaster-relief organizations have come to us and said, “This would be really exciting if you could put it into this platform because of these reasons.” So we’re seeing new customers and use cases emerge.

James (Host): Great. Yeah, I imagine disaster relief would be a great one—very applicable there. You have to have something that can withstand that environment.

James Thomas: Yes. With disaster relief, we’ve been looking at a lot of recent issues related to climate change. Whether people accept the term globally or not is a different matter, but if you look at what’s gone on with Californian wildfires, for example, we think it’s really important to help people in that situation and put things back together.

We’re expecting more of those kinds of events. When we look at that, it’s not just about day zero when it starts. At the beginning you’ve still got fiber to a lot of houses, you can contact people. By day one, you’ve lost chunks of fiber to some houses but not others. How do you communicate as a first responder?

Having a private network means that the bandwidth used by people panicking and calling loved ones isn’t shared with first responders—they have their own private slice.

Then there’s the recovery phase. If it’s all gone really badly—whatever the disaster is—now we need to talk about rebuilding and recovery. You’ve lost all your fixed infrastructure, so how do you coordinate the cleanup? For something of that scale, it’s huge. It’s not a one-person situation; we’re talking hundreds of people needed to clean up. When you have that many people, they need coordination and situational awareness.

As a society we’ve become so reliant on data for those things that the use cases we see are rich and varied across the whole spectrum. If I ask what you used your phone for today, it’s probably texting a loved one, going to a meeting, sending a work email without logging into your laptop, navigating somewhere, and so on. It’s about making sure those functions still happen.

James (Host): Yeah, I have a friend who lives in the South, and the hurricane that came through last year left him without internet for about 12 days. We couldn’t contact him, we couldn’t check if he was okay, nothing. Something like what you’re describing would’ve been fantastic—even just enough bandwidth for him to say, “Hey guys, I’m all right.” Once it goes down, it’s down.

James Thomas: Exactly. We’re running a project in Scotland for this very reason right now. If it’s gone down, what do we do about it? How do we solve this problem?

It’s not simple because there’s no one-size-fits-all. Not everyone has the same needs. You can’t just say, “We’ll throw some Starlinks at it.” Yes, you can put a Starlink terminal on the ground and suddenly there’s a Wi-Fi bubble there, but that doesn’t mean the house a mile away can connect.

How does a business function? You can share a Starlink connection if it’s properly sliced and give everyone a small amount of internet—back to dial-up days. You can at least do your emails, but maybe not join video meetings. How do we manage that?

We see real value in offering a solution that not only gives you a comms backhaul or link, but gives you a way to manage your services. A lot of that is around engaging with businesses and asking, “What are the absolutely essential functions? How much capacity do you need to get through a period like this?”

Two weeks is actually a long time. People do all their shopping online, contact family online, pay bills online. You need to renew your car insurance—how do you do that without internet?

James (Host): Yeah, it’s crazy. We’ve become so reliant on this stuff that it’s hard to function without internet access.

James Thomas: Exactly. And we’re seeing the same at sea, which is where the other side of what we do comes in. People working at sea will say, “Great, I’ve got Starlink on my vessel, but the second I leave my vessel I need direct-to-cell service, otherwise I don’t have comms.”

How do they get reliable comms with decent bandwidth—not just a text message—so they can do more interesting things? One of the projects we’ve looked at is using drones to map wind turbine blades. How do you do that without a decent comms link?

Do you want to do that in real time, or in arrears where you collect the data, go back, and only then realize, “There might be a crack there, so let’s send someone back out”? It’s those kinds of use cases we focus on.

James (Host): That actually brings me to a point I wanted to discuss: AI and the role of AI in what you do. In the drone example, if you’re scanning wind turbines and reporting that data, there’s probably an algorithm or AI or machine learning doing that. How crucial is this technology?

James Thomas: I think we need to be clear as a society about what is AI and what is ML, and where we sit on that spectrum of use cases. A lot of the most applicable stuff we’re doing right now is probably in the ML space.

It’s relatively simple logic such as: if you see a crack, take a better photo and then report it back so we can look at it properly later, rather than asking the system to learn whether that will cause a failure. The more advanced AI will follow, but it depends on what we’re actually trying to achieve.

If you look at building a true AI algorithm from scratch, you ask: is the training dataset good enough? What are we asking it to do? We’re not asking it to build future training sets by itself just yet. There’s a line in the sand around what AI actually gives us in practice.

James (Host): Yeah, I think you’re right, because people often use “AI” as a catch-all term for anything advanced these days. You really do need to break it down into those categories and decide what you need for your particular product.

James Thomas: Yes. We ask: are we actually building a whole new network from scratch? What are we doing with it? How are we getting to the answer it gives?

We did some really cool projects with RF data. We took raw RF data and processed it in snapshots using some GPUs, scanning gigabytes of spectrum data per second. We were trying to understand what signals were doing what. You’d never want to analyze that by hand.

That’s a great example where we saw it as ML: learning what a signal is over time. We had to teach it, but it wasn’t an all-singing, all-dancing “intelligence” in the sci-fi sense. Still, it’s going to be interesting to see how that plays out.

James (Host): I’ve spoken to a few guests on the show about this, and a common conclusion is that the biggest thing AI has given us is the ability to analyze and break down the data silos we’ve accumulated over the past few decades. For a long time it was just, “Let’s collect data,” and then, “What do we do with it?”

How do we analyze millions and millions of gigabytes of data? That’s where AI really comes in. It saves us time so we can focus on the more complicated design aspects while it breaks down the information.

James Thomas: I used it to build some software last night. I was playing around with something. I got a bit bored—it was 10 p.m., my partner was away—so I thought, “I’ll be an engineer again rather than the CEO for an hour or two.”

I was trying to program an Arduino—simple, off-the-shelf, inexpensive kit. I got AI to build some code for me, compiled it, tried it, then thought, “I could put these chips on it and add this as well,” so I had it generate that function set and added it. That worked. Then I wondered what would happen if I wanted to run a screen off it and move from an Arduino to a Raspberry Pi. I had a Pi lying around, so I tried that.

We use two or three different tools—ChatGPT, Gemini, and a few others. It’s interesting: if you run code through one, then ask another model to check the homework, you end up with a codebase that’s been checked by a different language model. Put it through two or three times, load it all up, and it works pretty well.

By the end of the evening, I’d effectively written an app with eight reference libraries, some new driver files for new chipsets, and tied it all together. Now all I have to do is build a PCB instead of using a breadboard.

That was so much quicker than even three years ago. In the past it would’ve taken me ages. I’m not a brilliant programmer—I used to write code to raise and lower my chicken-coop door, that was my level. So for simpler programming tasks it has made a massive impact.

James (Host): Yeah, it’s a tool. People are beginning to see it as, “This isn’t going to end my career; this is something that will save me time if I use it properly.”

James Thomas: The key is “use it properly.” You still have to put the right prompt in and know what “good” actually looks like.

James (Host): Yes. It’s not something where anyone with no understanding can just type in a prompt and magically get working code. You still need to understand what you’re creating.

James Thomas: Exactly. It’ll be interesting to see what it does to skill sets. I was born in the late ’80s, so I learned to drive in a manual car with nothing automatic. If you wanted the wipers, you moved the stalk; you wanted the lights, you turned the knob.

Now you get in a car and the lights come on automatically. You see a surprising number of people driving without lights because they don’t understand what’s happening. I wonder what AI will do in similar ways, where people say, “I don’t normally do this myself; I don’t understand the underlying science.”

I think it’s critical we remember what the inputs are and what “good” looks like. If you’re driving at night, you want the proper headlights on, not just sidelights, because the computer can’t always make the right decision.

James (Host): Right, it’s coming, it’ll just take time. So, as far as scalability goes, are there other sectors outside of what we’ve discussed where you’re planning to use your technology?

James Thomas: I think our technology is very good as a dual-use technology. It can be taken across a couple of different markets. I’m slightly careful answering this because in our journey we’re trying to target the novel and niche areas that no one is really playing in.

We get our own pool of really interesting work and can stay involved in the fun stuff. We’ve realized those “novel and niche” pools are actually quite big. There are a lot of big niches out there. I’ve enjoyed that space, and I think it could still get us to a unicorn-scale business in due course with plenty of revenue. It’s just that we’ve done the really hard stuff, which has been a lot more fun to solve than putting radios on masts in the middle of town centers. That’s not our bag.

James (Host): What about the space of developing nations—helping communities that don’t have any access to the internet? Is that something you’ve thought about using the technology for?

James Thomas: Yes. I’m involved in another startup I’ve invested in where we’re doing something pretty interesting. We’re using a waterfall in Nepal—in the middle of the Himalayas—to generate electricity with a generator set. We use that electricity to power an electrolyzer and generate green hydrogen.

We then pipe that hydrogen into houses so they can cook with a non-fossil fuel. So we have clean energy that’s waterfall-generated in a valley in Nepal. It took the guy who went out there three days to get there from London—it’s that remote.

You look at communities like that and there’s absolutely an opportunity to provide them with pop-up comms so they can connect to the rest of the world. They could be digitally connected to the UK in seconds, but at present it’s a three-day physical journey. From a job-creation point of view, they could do so much more productively if they had connectivity.

James (Host): I remember reading that the United Nations said access to the internet is now as influential in alleviating poverty as access to education, which makes sense when you think about it. The amount of knowledge, training, and opportunity you can access online is astronomical.

James Thomas: Exactly. You can teach yourself pretty much anything these days. I’ve tried to teach myself guitar on YouTube for ten years and failed every time, so you can probably teach yourself anything!

James (Host): (laughs) Yeah, I’ve never succeeded there either. But your technology’s capabilities are really exciting—you can genuinely help people.

James Thomas: I think so. The nice thing is being able to make a difference as well as doing it for the right reasons. First responders, disaster relief—that sort of use case is really valuable.

I’d also like to look at the mining industry. I’m not inherently pro- or anti-mining, but making things safer underground is an important potential use of comms. Doing underground comms and similar things is not a traditional place to build a 5G network, but it could be impactful.

James (Host): Right. Oil rigs too, I imagine.

James Thomas: They’ve spoken to us. I think there’s definitely market opportunity there, but I don’t think we’d want to prime that necessarily.

We’ve got impact investors. We want to create a positive net impact in what we do. We’re primarily focused on offshore wind. We need raw materials, but let’s try to extract them more efficiently and sustainably.

I’ve had quite a few long discussions with our investors: we want to create a positive impact, and we need to be a big business to do that. People are going to buy this technology anyway, so we want to support that, become big enough, and then have the positive impact—buying our metal from renewable sources and doing things the right way. It’s sometimes quite hard to find that line, though.

James (Host): Definitely. Are there any particular projects at the company that you wanted to highlight—anything you’re really excited about or proud of?

James Thomas: I’m really excited by quite a few things right now, but not all of them are public yet, so unfortunately I can’t share everything. There’s a very exciting project we’re doing with a utility—but maybe next time.

James (Host): Sure, no problem. And I know your wider company does a lot of training in the engineering space. Is that something you’re still doing these days?

James Thomas: We do some training, mainly on the use of our products now. That’s the main focus. We’ve tried to shift ourselves more toward being a product-based business than a consultancy, and I think that’s been a good move—more product-focused.

James (Host): Definitely. How have you personally seen the engineering space change over the past decade or two?

James Thomas: Massively—absolutely massively. Twenty years ago when I was starting my career it was very manual. There were limited amounts of computer processing being used for things. Drawings were still quite manual.

I remember working at Yazaki for Ford on a multi-billion-dollar project. I was using a printout and a red pen in around 2012–2014. We were still at the point where we were scanning things and using red pen and ink. I can’t imagine doing that now. People would think I was mad if I said, “Let’s print this 50-page drawing set and mark it up by hand.”

I used to quite like doing that at night, when everyone had gone home—laying out the entire harness drawing for a vehicle across a bench, 30 pages in a row, and marking it up with a big red pen. But not anymore.

Going forward, the amount of tech we use now is huge. Our graduates are doing really exciting stuff—all their modeling and simulation is done with modern tools, all the analysis is done with software.

We’re also starting to see AI used in building results and reporting. Instead of writing a test plan from scratch, why not paste results into a model and say, “Give us the conclusions, what does the report say?” It spits out a draft far more quickly.

James (Host): Yeah. I imagine the speed of everything has changed significantly—the whole process has been shortened.

James Thomas: Definitely. The design lifecycle is much quicker, so it’s much faster to go from concept to reality.

As we discussed with AI: what I did in an evening would have taken a decent programmer two or three months not that long ago. You still have to know which prompts to use, which chip you want to drive, some basics about what you’re trying to achieve. But in an evening, I did what would have been a sizable project previously.

James (Host): Yeah, and like I mentioned earlier, it lets you focus on the actual design aspects rather than the grunt work of programming, which is nice for you as an engineer.

James Thomas: Exactly. You’re more of an architect and tester rather than purely a doer. You set the problem and direction, but don’t have to manually write every line of code—the system does that and you validate it.

James (Host): And to build on that: are there any recent trends, maybe in the past 6–12 months, that you’ve noticed in the engineering space that have caught your eye?

James Thomas: We’re definitely seeing more awareness of supply-chain vulnerability and more action around mitigating it.

We’ve just seen what’s happened to Nvidia’s stock over the last few days. A Chinese company has basically built something comparable to what a pretty big Silicon Valley firm has spent a lot of money and time developing—and they did it quickly.

We’re starting to see how fast it is to build products now, and the next one will be even quicker. So we ask: what’s the impact of that, and how do we build supply-chain resiliency, especially in the digital space? The digital space relies on hardware—CPUs, GPUs, and so on. Silicon will be really valuable going forward in terms of resiliency.

How do we buy silicon, make PCBs, and ensure we have sovereign capability rather than relying on third-party countries where relations might be hostile, while still staying ahead in this data revolution?

James (Host): Yeah, I definitely think COVID changed the entire world’s viewpoint on supply-chain security. It used to be all about outsourcing, and now people realize, “Maybe it wasn’t a great idea to move everything to one country outside where we operate.”

James Thomas: Exactly. When I was much younger, I had a little company hauling gear for rock bands. We didn’t make much money, but we went on a lot of tours and had a great time.

I remember a guy in Luton who tried to make knobs for us—the little volume knobs for amplifiers. He was charging about two pounds each, so I outsourced them to China. Two days later I had a photo: “We’ve manufactured one, is this right before we send you the whole box?” They were about three pence each.

It was so much cheaper. At the time it was a clear decision. But prices are going up there, and I think tariffs are going to be interesting as well—how tariffs play out across all of this.

James (Host): Yeah, the past year and a half has been interesting. No one’s really secure enough in what they know to make firm predictions, because every time people think, “This is how it’s going to play out,” something happens—Ukraine, the Suez Canal blockage, Chinese factory shutdowns—things out of our control that change the entire prediction model.

James Thomas: It’s been a very volatile few years. It makes you realize the period of stability we had until about four years ago. Now you’ve got geopolitical tensions in the East, all the things you mentioned, and big political changes.

You have a new US president, which will affect market forces, and in the UK we’ve gone a different direction. Eastern Europe is in a changing situation. It’ll be interesting to see what that does to our sector.

It will drive supply-chain resiliency one way or another, depending on how it plays out. It’ll drive where money is spent—defense, energy, oil and gas, and so on. There will be a lot of change.

James (Host): Right. And you also have emerging manufacturing hubs that weren’t really considered before—Mexico, Vietnam, Malaysia, Indonesia—now in focus as places with affordable labor and good quality manufacturing.

James Thomas: Exactly. Countries not traditionally thought of as manufacturing powerhouses, but safer and more trustworthy for IP. The question is: are they capitalized enough, do they have the ability to deliver, can they deliver on time?

There’s a huge opportunity there. If we send work to Mexico from the States, that’s probably a fairly safe vendor compared to China. You can drive across the border, it’s land transport, you know what’s in the box, you can visit relatively easily.

We’ve done similar things in the UK with Eastern Europe—many of our manufacturing partnerships have been based on that logic.

James (Host): I know Eastern Europe is one of the biggest vinyl manufacturing hubs in the world for music—vinyl records. Every person I know who owns a record label in the UK manufactures in Eastern Europe.

James Thomas: Really? That’s pretty cool.

James (Host): I think since Brexit it’s become slightly more expensive, but still cheaper than many alternatives.

James Thomas: Brexit—what a gem we lived through.

James (Host): (laughs) So that brings us to time. I just want to ask one last thing. Is there anything coming up for Jet that we can be excited about or keep an eye out for?

James Thomas: We’re getting some of our big offshore platforms into the water very shortly for their first major customer deployments, and that’s going to be super exciting.

They’re 33-ton ocean data hubs, analyzing data in real time, performing AI at the edge in real time, and transmitting that back to shore. We’re running a 5G network there in real time, and we’ve got floor-to-ceiling server racks floating continuously, 24/7.

In terms of big data, we can harvest raw video feeds continuously, transmit them 24/7, or do a lot of edge-processing AI. It’s going to be amazing to see what we do with all that data. We can harvest so much, and these platforms will be fully instrumented for one of our big customers. I’m excited to see what comes out of that.

James (Host): Awesome. Really exciting. Again, James, thank you so much for coming on. It’s been great chatting with you. Before we wrap up, just to help you get the word out: if anyone wants to follow the company, support the company, or stay up to date, what are the best places to do that?

James Thomas: Our website is www.jet-eng.com. People are also welcome to connect with me on LinkedIn—James Thomas—and I’ll happily have a conversation if anyone wants to follow up.

James (Host): Fantastic. Well, thank you again, James. It’s been great chatting with you, and I have a much better understanding now of maritime technology. It’s really complicated stuff.