This Vectis push looks bold. It also looks like the kind of program that makes people fall in love with the slide deck and forget the part where the real world fights back. Building four more UAVs and aiming for a first flight in 2027 is a serious signal, not a hobby project. But the more “modular” and “integrated” a drone is supposed to be, the more brittle it can become when the shooting starts and everyone’s assumptions get stress-tested at once.
From what’s been shared publicly, Skunk Works has expanded the Vectis UAV effort by producing four additional aircraft. The stated plan is still a first flight in 2027. They’re framing Vectis as survivable and adaptable in contested environments, and they’re leaning hard on two themes: speed and evolution. Concept to wind-tunnel testing in under two months, digital model in six, a first UAV in eight. Integration with manned aircraft like the F-35 is treated as a core feature, with the promise that it extends range, unlocks new tactics, and keeps pace with threats.
As a company that lives on the other side of that promise—building drone detection radar systems and AI that fuses inputs from different sensors—this kind of announcement lands in a very specific way. Because the story is not just “new drone.” The story is “new drone designed to be a moving part in a larger system.” That’s where things get exciting, and that’s where they get dangerous.
If Vectis really is modular in the way they’re implying, then the airframe is almost the least interesting part. The real product is the ability to swap sensors, change payloads, update tactics, and keep the same platform relevant as missions shift. That’s smart. It’s also a quiet admission that nobody knows what the next few years of air combat will demand, so they’re building something that can be rewritten.
Here’s the tension: “evolvable” is a great word until you ask who is responsible for the evolution. If the drone keeps changing, then detection, identification, and tracking have to keep changing too. That is not a moral complaint. It’s a practical one. Every new payload, every new flight profile, every new emission pattern becomes a new set of problems for radar drone detection and for the sensor fusion that turns raw signals into a confident track.
And when the UAV is meant to work alongside an F-35, that pressure multiplies. The whole point of pairing unmanned and manned platforms is to stretch the formation—push capability outward without putting a pilot in the riskiest place. Imagine a scenario where a crew depends on the unmanned wingman to take the first look or carry a payload farther forward. If the UAV is detected earlier than expected, or tracked more cleanly than expected, the “extra flexibility” turns into a new vulnerability. Suddenly the manned aircraft is reacting to a problem it thought it had outsourced.
The speed claims are what I’m most conflicted about. Moving from concept to prototypes quickly can be a real advantage. Iteration is how you learn. But speed also has a way of hiding risk because you can always say, “We’ll fix it in the next build.” That works fine for some things. It works less fine for survivability, signatures, and the messy physics of being seen and tracked.
And yes, survivability is not just armor and countermeasures. It’s also whether the drone behaves in ways that don’t scream “here I am” across radar, passive sensors, and whatever else is watching. In the real world, defenders don’t rely on one sensor. They stack them. They cross-check. They get suspicious when one channel goes quiet. That’s why we invest so heavily in fusing different sensors into one picture—because attackers adapt, and defenders adapt right back.
There’s also an uncomfortable political reality: “for the United States and its allies” means this won’t stay contained. If Vectis succeeds, it shapes expectations. Partners will want similar capabilities, similar integration, similar timelines. That can be good for deterrence. It can also accelerate the pace at which contested airspace becomes crowded with semi-autonomous systems on all sides, each one trying to out-think detection and each defender trying to prove they can still see what matters.
A fair counterpoint is that modularity might actually reduce risk. If you can update the drone as threats change, you’re not stuck with a platform that’s obsolete the day it flies. And integrating with aircraft like the F-35 could reduce decision time and reduce mistakes, because the manned platform becomes the coordinator and the unmanned platform becomes the tool. That’s a compelling vision.
But it depends on details that nobody can hand-wave away: the quality of the integration, the discipline around updates, how testing is done, and how often real detection systems surprise the designers. In our world, surprises are normal. A drone that looks “survivable” in one test range can look very different when multiple sensors, multiple angles, and real tactics are involved.
So yes, building four more aircraft matters. It suggests confidence and momentum. It also raises the stakes for everyone—builders, operators, and defenders—because the second-order effect of faster iteration is faster confrontation between new drones and new ways to find them.
If programs like Vectis keep accelerating, what standard should decide whether “modular and integrated” is actually safer and smarter in combat—or just a faster way to scale risk?