This is the part of the war that should make anyone selling drone defense tech a little uneasy — and a lot more serious. Not because drones are “winning” on their own, but because a cheap act on the ground can suddenly make expensive protection feel like a costume. If a few people with access and intent can knock out the right link in the chain, the sky gets dangerous fast.
Based on what’s been shared publicly, a Russian partisan brief covering 31 May to 07 Jun 2026 points to a pattern: sabotage aimed at military infrastructure and at the war economy, plus open recruitment messages trying to pull more insiders into resistance. The headline examples are blunt. One group, ATESH, reportedly sabotaged cellular towers near Perm. The claim is that this disabled Russian electronic warfare systems in the area and helped enable Ukrainian drone strikes on major oil and chemical facilities. Another group, Black Spark, reportedly used a remote detonation to hit a diesel tank car in Krasnodar — framed as targeting “oligarchic” infrastructure connected to the war economy, and contributing to fuel supply disruption in southern Russia. A third, the Kuban Partisan Detachment, put out a recruitment appeal aimed directly at civil servants and military personnel, basically telling them: you’re part of the machine, and you don’t have to be.
If you build radar drone detection and sensor fusion systems like we do, you read this and you don’t just think “security problem.” You think “systems problem.” Because what’s being attacked here isn’t only a target. It’s reliability. It’s continuity. It’s the invisible support structure that lets air defenses, communications, jamming, and response teams function as one coordinated thing.
And I’m going to say the uncomfortable part out loud: a lot of drone defense conversations are still too obsessed with the object in the air and not obsessed enough with the ecosystem underneath it. People ask, “Can you detect a drone?” Sure. But detection is a promise you have to keep during chaos. If local communications go dark, if power is unstable, if teams can’t coordinate, if someone inside a maintenance chain decides to “forget” something important, then your shiny stack turns into a set of disconnected boxes.
The ATESH example is especially telling because it’s not even a direct strike on a radar site or a command post. It’s an attack on the surrounding tissue. If the report is accurate, knocking out cellular towers near Perm hurt electronic warfare capability — and that created a window for drones to get through to oil and chemical facilities. Even if the details are messier than the brief suggests (they usually are), the lesson stands: you don’t always defeat air defense by overpowering it. You defeat it by making it blind, slow, or confused at the exact wrong time.
Now zoom out to the Black Spark claim about the diesel tank car. This isn’t “battlefield” sabotage. This is pressure on fuel and logistics, which then pressures readiness, movement, repair, and morale. A drone can’t be intercepted if the response vehicle isn’t fueled. A site can’t run if backups don’t get delivered. You can have excellent radar drone detection on paper and still lose the practical ability to act.
And then there’s the recruitment appeal, which is a different kind of strike. It’s not against hardware. It’s against loyalty. If you’re trying to protect infrastructure, internal dissent changes your threat model in a way most organizations hate to admit. Not just because of espionage, but because of passive resistance: slower repairs, missed checks, conveniently delayed deliveries, a gate left open, a camera pointed the wrong way. That’s not a movie plot. That’s how real systems fail when people stop believing in the mission.
This is where our own industry can be a little too comfortable. We like to talk about performance and ranges and clever fusion. We should. But we also need to talk about resilience: What happens when one sensor goes down? What happens when your network is degraded? What happens when you can’t trust every node? Fusion only helps if the inputs stay available and believable. If sabotage corrupts the inputs, you can fuse yourself into confidence you didn’t earn.
Imagine you’re a security manager at a refinery that might be targeted. You install a radar drone detection layer, some optical sensors, maybe acoustic, and you feel good. Then one night the local comms are disrupted, your handhelds don’t work, your team can’t confirm tracks, and your response time doubles. The drone didn’t get smarter. Your environment got weaker. That’s the real risk: defenders keep upgrading the “eyes,” while attackers keep targeting the “nerves.”
To be fair, there’s another read on all this: partisan claims can be exaggerated, and linking a tower sabotage to a specific strike might be more storytelling than proven cause. Also, if sabotage increases, it could lead to harsher internal controls, tighter guard forces, more redundancy, and faster repair — which might reduce the payoff over time. But that doesn’t make the trend harmless. It just means the contest becomes more brutal and more paranoid, and critical infrastructure sits in the middle.
So here’s the point we’d argue, even if it annoys people: drone defense isn’t mainly a product problem anymore. It’s an operational trust problem. If your plan assumes stable communications, loyal staffing, and clean logistics, you don’t have a plan — you have a wish.
If partisan activity keeps expanding from physical sabotage into recruitment and insider influence, what should matter more in protecting sites: better detection tech, or better resilience against the human and network weak points that make detection useless?