This is the kind of headline that looks clean and simple—“stock up on a new order”—but the real story is messier, and higher-stakes. A $10 million demining order sounds like a win you take and move on. If you build drone detection radar systems and AI that fuses different sensors, you don’t get that luxury. Demining isn’t a “nice project.” It’s a test of whether autonomy can be trusted around life-or-death decisions, in places where mistakes don’t get patched later.
Based on what’s been shared publicly, Ondas is up 6.4% after an initial $10 million order tied to an autonomous demining program in Israel. The same public reporting says this pushes their aggregate active tender value to $80 million. They also recently launched a joint venture, ONBERG, with Heidelberg, focused on counter-drone defense systems in Germany and Ukraine.
Those are the facts. Here’s my read, from the company side of the table: this is Ondas telling the market, “We’re not a single-product defense bet. We’re building a full autonomy stack across threats.” Demining on the ground, counter-drone in the air, and a strategy that says the software brain—sensor fusion, decision support, autonomy—can travel across missions. That’s ambitious. It’s also where companies like ours either become essential infrastructure or become a cautionary tale.
Because the hard part isn’t winning an order. The hard part is delivering autonomy that soldiers and operators will actually trust when stress is high and information is ugly.
Imagine you’re a unit commander. You’re told an autonomous demining system cleared a route. You’re looking at the clock, you’re looking at your people, and you’re looking at a piece of technology that you didn’t build. If the system gets it wrong, you don’t get a refund. You get casualties. That’s the standard we’re playing under, and it changes how you should read “initial order.” It’s not just revenue; it’s a trial run in credibility.
Now zoom out. A $10 million initial order can be two things. It can be a foothold that expands fast if performance is solid. Or it can be a contained experiment that never becomes a scaled program because the last 10% of reliability is brutally expensive. People love to talk about autonomy like it’s a straight line—more data, better model, problem solved. In defense, the curve isn’t smooth. It’s cliffs. One bad incident can freeze adoption for years, not only for one company, but for the whole category.
This is where radar drone detection and multi-sensor fusion matter, even if the headline is “demining.” Modern battlefields are layered. A demining operation isn’t happening in a vacuum. Drones watch. Signals get jammed. Weather changes. Civilians move. If your autonomy stack can’t keep operating when one sensor lies, or one channel drops, or the enemy adapts, you don’t have a product—you have a demo. And markets are finally learning the difference.
The ONBERG joint venture angle adds another layer of tension. Working across Germany and Ukraine isn’t just “expanding footprint.” It’s building in two environments with very different rules, oversight, and pace. In Ukraine, the pressure is speed and survival. In Germany, the pressure is controls and accountability. If you’re serious, you have to satisfy both. Do that well and you build a reputation for systems that work and can be responsibly deployed. Do it poorly and you get the worst combo: slow sales, high scrutiny, and a product roadmap driven by politics instead of reality.
There’s also a temptation here that I think is dangerous: treating tenders like victories. An “aggregate active tender value” of $80 million sounds impressive, and it might be. But tender value is not cash in the bank, and it’s not proof of field performance. If leadership starts managing the company to the tender scoreboard, you can end up optimizing for proposals instead of outcomes. That’s how defense tech companies lose their way—busy, visible, always “in the mix,” but not reliably deployed at scale.
To be fair, there’s a strong counter-argument: autonomy is exactly what reduces risk for humans. If machines can clear mines, humans don’t have to. If counter-drone systems can detect and stop threats early, fewer people die. I agree with that. I’m in this industry because I think it’s directionally right. But I don’t think “directionally right” is enough. The moral weight of these systems means we need to be harder, not softer, on proof.
So yes, a 6.4% pop makes sense. Markets reward momentum. But inside companies like ours, the real question is whether this order forces the discipline we say we have: testing in harsh conditions, honest reporting on failure modes, and building operator trust over hype.
If you’re betting on autonomy in defense, what should matter more to you right now: the size of the tender pipeline, or the evidence that these systems can be trusted when the stakes are human lives?