Dual-use tech always sounds clean on a slide: safer infrastructure, smarter industry, stronger defense. In real life, it’s messier. And BGV launching a dual-use tech hub while backing an Israeli drone firm like Percepto is a perfect example of why this moment is both full of opportunity and full of traps.
From what’s been shared publicly, BGV is leaning into the intersection of AI, infrastructure, and defense, betting that demand is surging. Percepto builds fully autonomous inspection systems designed for harsh industrial environments. That’s the headline. The subtext is the real story: the line between “industrial automation” and “security capability” is not a line anymore. It’s a hallway with a lot of doors, and everyone is walking through them.
From where we sit—as a company that builds drone detection radar systems and AI that fuses signals from different sensors—this move makes sense. Not because it’s trendy, but because the world is forcing it. Industrial sites, energy assets, ports, rail yards, warehouses, even big construction projects: they’re all getting more automated, more connected, and more exposed. The same conditions that make autonomous inspection attractive also make these places easier to disrupt.
Here’s the uncomfortable part: a drone that can inspect a refinery safely is not far away, conceptually, from a drone that can probe where that refinery is weak. A system that can navigate “harsh environments” reliably is also a system that can operate where people can’t follow. Dual-use isn’t a category. It’s a reality that shows up the minute the tech works well.
So when a major investor puts real weight behind a hub like this, the consequence isn’t just “more innovation.” It’s acceleration. It’s faster buying cycles. It’s more pilots that become rollouts. And it’s more pressure on everyone in the ecosystem to answer questions they’d rather avoid: Who is the end user? What rules govern deployment? What happens when a “civilian” tool ends up shaping a security outcome?
If you build in this space, you can’t pretend those questions are someone else’s job.
We see it on the customer side. A plant manager might come to us saying, “I just need to protect my perimeter.” They’re not asking for geopolitics. They’re asking because they’ve had unexplained drone sightings near sensitive operations, or because their insurer is suddenly asking what they’re doing about aerial threats, or because a regulator is getting stricter after incidents somewhere else.
That’s where radar drone detection stops being a “defense thing” and becomes a basic safety and continuity thing. If a drone gets too close to a flare stack, a substation, or a loading area, it’s not only a security incident. It’s a shutdown. It’s a headline. It’s people getting hurt. The stakes are not abstract.
Now zoom out one level. If more investors fund companies like Percepto, you’ll see more autonomous systems in more places. That’s good—until it isn’t. Because every autonomous asset you add to a site creates new assumptions: that it will behave as expected, that it won’t be spoofed, that it won’t be hijacked, that it won’t become a blind spot. And when you have many autonomous systems operating at once, the only way to keep the picture clear is sensor fusion—radar, cameras, acoustic, network signals—pulled into one view that a human can trust under pressure.
That’s not a nice-to-have. It’s the difference between “we noticed something odd” and “we can prove what happened.” It’s also the difference between responding calmly and overreacting.
And yes, overreaction is a real risk here. One scenario we worry about: a site gets spooked by drone activity, rushes a half-tested detection system into production, and suddenly every bird, crane, or weather effect becomes an “alert.” The security team starts ignoring alarms. Then the day a real drone shows up, it blends into the noise. The company thinks it bought safety, but actually bought fatigue.
Another scenario: a city authority installs detection around key public sites. The tech works, but the governance doesn’t. Who gets access to the data? How long is it stored? What counts as a threat? You can build a system that detects drones reliably and still end up in a public trust crisis if people think the goalposts are moving.
To be fair, there’s a strong argument on the other side: dual-use investment can harden society. Better inspection means fewer industrial accidents. More automation means fewer people sent into dangerous spaces. Better detection means fewer disruptions. If geopolitical tensions are rising and critical infrastructure is under strain, waiting for “perfect clarity” can be its own kind of irresponsibility.
I get that. We build for the world that exists, not the world we wish we had.
But here’s my judgment: the biggest danger in this rush is pretending the technology is the hard part. It isn’t. The hard part is deciding what “normal” looks like when advanced tools become cheap and everywhere. The hard part is making sure the buyers understand what they’re buying, and that the sellers stay honest about what the systems can and can’t do. The hard part is building deployments that don’t break under real conditions—weather, clutter, spoofing attempts, human panic, and the simple chaos of a working site.
BGV’s move signals that dual-use is not a side bet anymore. It’s a strategy. That means companies like ours will see more demand, more partnerships, and more pressure to move fast. We should welcome the momentum—but we should also insist on clearer boundaries, better testing, and better accountability, because the downside of getting this wrong is not just wasted budget. It’s disrupted infrastructure, escalations based on bad data, and a world where trust collapses faster than technology improves.
If dual-use hubs keep scaling and autonomous systems spread into more critical places, what’s the one rule you think should never be compromised—speed, safety, transparency, or control?