In January 2025, as wildfires tore through Los Angeles, the cell networks in and around the fire areas went nearly dark. Independent network monitoring confirmed the collapse: service on all three major carriers cratered, bottoming out on January 8. Coordination between responders suffered. Some evacuation notices to people directly in the path of an advancing firestorm never went out, because the infrastructure that was supposed to carry them was on fire.

That was not a freak event. It is the pattern. During a 2025 wildfire outage in Oregon, emergency management had to deploy eight Starlink terminals just to keep basic connectivity alive. When the Maui fires hit in 2023, officials put it bluntly: “911 is down. Cell service is down. Phone service is down. That’s been part of the problem.” In the Kincade fire, 874 cell sites went offline, more than half the sites in one county. Hurricanes, floods, earthquakes, deliberate public-safety power shutoffs, the story is always the same. The disaster destroys the exact communications infrastructure that the response depends on, at the exact moment coordination matters most.

Sit with the cruelty of that timing. A shared common operating picture, everyone seeing the same map, is never more valuable than during a chaotic, fast-moving emergency. And that is precisely the moment the infrastructure to carry it tends to be gone.

The quiet assumption in every TAK server

TAK’s whole value is the shared picture, and the shared picture, past a couple of phones on the same Wi-Fi, needs a server. We said that in our last post: the app was never the barrier, the server was.

But there is a second, sneakier problem with the standard ways people run that server, and disasters expose it. Almost every normal TAK server deployment quietly assumes working infrastructure. A cloud server assumes the internet is up. A server in a building assumes that building, its power, and its network connection are all fine. A server someone administers remotely assumes there is a path to reach it.

Those are reasonable assumptions on a normal Tuesday. They are exactly the assumptions that fail in the scenarios where you most need the tool. When the towers fall, a cloud-hosted TAK server is just as unreachable as everything else. Your shared picture dies with the infrastructure, right when the picture would have saved the most trouble. You did everything right, you stood up a real server, and it evaporated the moment the network did.

A server that depends on the infrastructure surviving is not a server you can count on in a crisis. It is a server you can count on until you actually need it.

So the server carries its own everything

Here is the fix, and it is almost stubbornly simple: make the server depend on nothing but the phone it runs on.

Our Edge TAK Server does exactly that. It turns a single Android phone into a complete, self-contained TAK server. It runs its own certificate authority, handles secure encrypted CoT streaming and device enrollment itself, and runs as a proper background service so it keeps serving with the screen off. And critically, it brings its own network. The host phone puts up a hotspot, and that hotspot is the battlefield network. Teammates join it, scan one QR code, and their stock ATAK imports a full identity package and goes green in about a minute.

Notice what is not in that description. No cloud. No internet. No data center. No IT ticket. No dependence on a single cell tower staying upright. The entire backend, the CA, the server, and the network it runs on, is one phone in one pocket. If that phone has power, your team has a shared picture, full stop, whether or not a single piece of public infrastructure is still standing for miles.

That changes what the tool can do in the exact scenarios that break everything else:

It works with zero connectivity. Because the server hosts its own local network, the team’s shared map does not care that the towers are down. The picture lives on the phone and the phones around it. This is the capability the LA responders needed and did not have: a coordination layer that keeps working when the carriers do not.

It deploys in minutes, anywhere. There is nothing to provision ahead of time and nothing to reach across a network. You pull the phone out, start the server, bring up the hotspot, and onboard people with a QR code. That is the whole standup. It is built for rapid deployment because that is the situation it was designed for.

It is genuinely pocket-sized. The deployable network assets the carriers send to disasters, the cell towers on wheels and rapid-deploy units, are real and valuable, but they are trucks and trailers that have to physically arrive. This is a phone that is already there, in someone’s pocket, on day zero, hour zero.

It costs nothing to try. It is free on the Play Store. There is no reason a small team cannot have it staged and ready before the season starts.

And when the wider network does come back, you are not trapped on an island. You can bridge out to reach other teams or a larger picture. The difference is that you never depended on that link. Connectivity becomes a bonus you use when it exists, instead of a lifeline that strands you when it does not.

Not just for the worst day

The disaster case is the sharpest way to see why this matters, but the same properties pay off on ordinary days too. Any small team that does not want to run cloud infrastructure, does not have an IT shop, or just needs a shared picture for a few hours gets the same deal: a full TAK backend with none of the overhead. A two-person detail. A pop-up operation. A training exercise in a field with no bars of signal. The phone-as-server does not ask whether the situation is a catastrophe or a Tuesday. It just works with what you have, which is a phone.

We are honest about the edges of this. If you are a large agency coordinating hundreds of users across a region with dedicated staff, a full server program is the right tool, and you should run one. Edge TAK Server is for everyone underneath that, and for everyone whose worst-case scenario involves the infrastructure being gone. Which, if you work anywhere near disasters, austere environments, or contested comms, is not a hypothetical. It is the whole job.

If you want a shared picture that survives the day the towers do not, this is it, and it is free to go try right now. And if you want it tuned to your team’s exact deployment, or bridged into something you already run, that is a conversation we are always happy to have. Email hello@spicycorp.com, or book a call.