Everyone loves a “we built it in 84 days” story. It sounds like competence. It sounds like urgency. It sounds like the defense world finally moving at the speed the drone threat demands.
And I still don’t buy the victory lap.
From what’s been shared publicly, Saab said it developed the concept for a system called Loke in less than three months — 84 days — aimed at detecting and engaging small enemy drones. The Trackfire combat module is part of that system as the fire element. Those are the facts on the table. No drama needed. The drama is baked in.
Because “concept in 84 days” can mean two very different things. One version is impressive and real: a tight team, clear requirements, smart reuse of proven parts, and a design you can actually field. The other version is what the industry sometimes calls speed: a clean slide deck, a neat name, a couple of demos in controlled conditions, and a long walk from “concept” to something that holds up when drones come in messy swarms, in bad weather, with jamming, with clutter, with civilians nearby.
As a company that builds drone detection radar systems and AI fusion from different sensors, I’m not offended by the sprint. I’m worried people will confuse sprinting with arriving.
Small drones are not hard because they’re magical. They’re hard because they’re cheap, many, and inconsistent. One is plastic, one is foam, one is modified, one is flying low behind buildings, one is coming in at a weird angle, one is being used as bait. A lot of this is boring reality: false alarms, missed tracks, clutter, and the human being in the loop trying to decide fast without messing up.
That’s where radar drone detection helps, but it’s also where it can betray you if you treat it like a silver bullet. Radar gives you reach and persistence, and it doesn’t “get tired.” But radar on its own doesn’t solve identification and intent. It gives you tracks. Tracks become decisions. Decisions become consequences. If your system confidently tells an operator the wrong thing, speed stops being a virtue and starts being a liability.
The moment you add “and engage” — like Loke is described — the stakes jump. Detecting a drone is one kind of problem. Choosing to shoot at it is another. Now you’re not just managing performance, you’re managing risk: collateral damage, misidentification, escalation, and the simple reality that a lot of places you need protection are also places you can’t just open fire casually.
Imagine a port trying to keep commerce moving while staying safe. A radar drone detection layer flags something small and low over the water. Is it a threat, a hobby drone, a bird, a sensor glitch? If the system pushes the operator toward “shoot” too fast, you may stop a drone and start a crisis. If it pushes “wait” too long, you may lose the window. The quality of fusion — radar plus other sensors, tied together with logic that handles uncertainty — is the difference between a smart defense and a nervous trigger finger.
The optimistic read of Saab’s announcement is that the industry is finally behaving like the drone threat is a software-paced problem, not a decade-long procurement project. Good. The world changed. Waiting for perfect requirements and perfect tests is a luxury many customers don’t have.
But here’s the uncomfortable part: moving fast can also be a way to dodge accountability. When something is “just a concept,” everybody can clap and nobody has to answer the hard questions. What’s the false alarm rate in real clutter? How does it behave near airports, stadiums, or dense cities? What happens when the enemy intentionally tries to saturate the operator with cheap decoys? How does the system decide when it’s not sure?
And yes, Trackfire being part of the system matters. A weapon module is a serious commitment. It means the detection side can’t be “pretty good.” It has to be dependable, explainable enough for operators, and stable under stress. The system can’t just detect; it has to support decisions that are legally and morally heavier than “track on screen.”
There’s also a competitive pressure here that people don’t say out loud. When one big name says “84 days,” everyone else suddenly looks slow. That can push teams to cut the unglamorous work: sensor integration details, edge-case testing, operator workflow, training, and the boring but vital question of what happens when the network drops or the environment changes. Those are exactly the cracks small drones slip through.
To be fair, a concept sprint can be the right move if it’s paired with disciplined follow-through. Rapid concept work can lock the architecture early, make integration easier, and get something into user hands faster for feedback. We do this ourselves on the sensor and fusion side: build, test, learn, tighten, repeat. Speed is not the enemy. Unearned confidence is.
If Loke becomes a real, field-tough system, that’s good for everyone on the defensive side. But if the “84 days” headline becomes the standard people chase, we’re going to see more rushed systems that look strong in demos and buckle in the first confusing, noisy, real engagement — and the people who pay for that won’t be the ones who approved the concept slides.
So the real issue isn’t whether 84 days is possible; it’s what we’re rewarding as an industry: fast announcements, or systems that handle uncertainty without pushing operators into bad choices.
What do you think matters more in counter-drone systems right now: the ability to produce a working concept quickly, or the patience to prove it holds up in the ugly real world before anyone claims success?