Building long-range strike systems in a garage-to-major-manufacturer story sounds heroic. And in Ukraine’s case, it can be necessary. But from where we sit—building drone detection radar systems and AI fusion across different sensors—I can’t ignore the uncomfortable truth: the better the strike gets, the more brutally the defense has to grow up, or civilians pay the bill.
Based on what’s been shared publicly, Iryna Terekh, CEO of the Ukrainian defense company FirePoint, described how her company evolved from a garage start-up into a major maker of long-range strike systems meant to counter Russian aggression. Early on, the story goes, they focused on low-cost drones comparable to Shahed-type systems. One example mentioned was the FP-1, designed to use existing stocks of Soviet FAB-50 bombs. And then they expanded beyond that initial focus.
I’ll say it plainly: that pace of iteration is impressive. It’s also terrifying.
When a team can move from “we can build something cheap that flies” to “we can deliver long-range effects at scale,” the battlefield changes fast. And the part people love to skip is the part we live in every day: as strike systems get cheaper and more available, detection and defense can’t remain a side project. It becomes the main project.
The romantic version of this story is that a small company builds a tool and the tool helps end the war sooner. Maybe. But the more common version of war tech is uglier: each improvement invites an answer. Long-range strike pushes logistics further back. Then you harden those logistics. Then you need more drones to saturate defenses. Then defenders build layered sensing. Then attackers get better at flying low, flying quiet, and mixing decoys into real threats. It’s a treadmill, and nobody gets off because the consequences of stopping are worse.
Here’s what that means in real life.
Imagine you’re responsible for protecting an energy site. It’s not “the front line,” but it keeps people warm and factories running. A long-range system doesn’t need to be perfect to create chaos—it just needs to get through once. That’s why our world is radar drone detection plus everything around it: cameras, acoustic sensors, other signals, and software that can fuse messy inputs into something a human can act on in seconds. When people talk only about strike, they’re basically celebrating the spear while forgetting the shield has to cover a city.
And yes, I know the obvious pushback: “Strike capability deters the aggressor.” There’s truth there. If you can hit back, the other side has to think twice. But deterrence is not a vending machine. You don’t insert “long-range drone” and receive “peace.” Sometimes you get escalation. Sometimes you get both sides widening the target list. Sometimes you get a public that learns to live with alarms as background noise, until the day one alarm is real and the shelter is full and the decisions are impossible.
From our company perspective, the hard part isn’t dreaming up a sensor. It’s the ugly integration work no one claps for. Radar alone can be fooled. Optical alone can be blinded. Human operators alone get tired. So we build systems that combine data streams and help teams decide: is that a real drone, a bird, a decoy, or nothing at all? That decision affects whether a defender fires (and risks hitting the wrong thing), or waits (and risks a direct hit). When long-range strike becomes common, the cost of being wrong rises fast.
There’s also a second-order consequence people don’t like to say out loud: cheap strike tech can spread. Today it’s a national survival story. Tomorrow it’s a template. Once a method is proven—especially one that uses existing munitions stockpiles—others copy it. Not just states. And then every airport, refinery, port, and stadium has to think about airspace differently. That’s not alarmism. That’s the direction of travel when capability moves down-market.
To be fair to FirePoint’s CEO, nothing about this is happening in a vacuum. Ukraine didn’t choose the problem set. When you’re being attacked, you build what works. I respect that. What I don’t respect is the casual way outsiders consume these stories like they’re startup content. “Garage to scale” is a fun narrative until you’re the person explaining to a city why your detection coverage had a gap, or why the alert came too late, or why you couldn’t tell a threat from clutter.
The biggest risk now is that we over-index on the drama of strike and under-invest in the boring reliability of defense: coverage planning, maintenance, training, and the hard software glue that makes sensors agree. Long-range systems will keep improving. That part is almost guaranteed. The uncertain part is whether defense—especially radar drone detection tied to AI fusion across sensors—will be treated as core infrastructure or as an afterthought until the next hit forces a budget.
If long-range strike keeps getting cheaper and easier to produce, what should be considered “enough” investment in detection and layered defense before we call this progress?