Where this came from
My fascination with drones started years ago when I bought my first drone: a first-generation DJI Mavic Pro. At the time, it felt like science fiction. For the first time, I could launch a flying camera from my backpack, send it hundreds of feet into the air, and see the world from a completely different perspective. What began as a tool for content creation quickly became something much more interesting to me.
I spent hours flying it simply because it was fun. Every flight felt like a glimpse into the future. The technology was already impressive, but what fascinated me most wasn't what the drone could do in that moment — it was what it might become. I found myself constantly watching videos of new drone developments, especially FPV pilots who were pushing the limits of speed, precision, and control. There was something incredible about seeing a machine become an extension of a person's reflexes and decision-making.
Eventually, I gave the drone away and moved on to other things. But a few years later, Russia invaded Ukraine, and the world got its first real glimpse of how rapidly drone technology was evolving. Suddenly, drones weren't just tools for photography or hobbyists. They had become a defining feature of modern warfare, reconnaissance, logistics, and asymmetric conflict. What had once seemed futuristic was becoming reality.
That moment brought me back to the same question I'd been asking years earlier: where is this technology headed?
I believe drones — and eventually autonomous drone swarms — are going to fundamentally reshape how nations think about defense, deterrence, and conflict. The pace of innovation is staggering, and it's only accelerating.
Enjambre is my attempt to contribute to that future.
The vision is to build a platform that allows people to design, simulate, test, and rapidly iterate on drone systems here in the United States — and ultimately turn that process into a competitive sport.
I grew up playing sports and briefly played college football. In my experience, nothing develops skill, discipline, and focus faster than competition. If swarm-on-swarm drone operations are going to become a defining capability of the coming decades, we'll need talented pilots, engineers, operators, and strategists who can make decisions under pressure and think clearly about resource allocation at scale.
This technology isn't slowing down. If anything, it's accelerating.
So why not build an arena where people can learn, compete, innovate, and have fun — while developing skills that could someday matter far beyond the game itself?
The problem
Everyone who has tried to let people fly real hardware from their couch has failed — Surrogate.tv pivoted, vRotors stalled, FlyThere went quiet. The dream runs straight into physics: glass-to-glass latency of 130–230 ms across the country makes raw, hands-on FPV flight literally unstable. On top of that, the FAA's visual-line-of-sight rules make "a customer in another state pilots a drone over open land" flatly illegal, and the RF spectrum tops out at roughly eight simultaneous pilots per cell.
The approach
Enjambre takes the failures seriously and reframes the problem around the three hard constraints. Latency: the flight controller stabilizes locally in GUIDED mode; the remote player sends velocity intent — heading, altitude, speed — instead of raw stick inputs, so the network only has to close a ~20 Hz loop. Legality: the whole venue lives inside a roofed, netted enclosure, which takes it out of the national airspace and out of the FAA's jurisdiction. Spectrum: several RF-isolated flight cells are stitched into one tournament by a server-side referee that scores hits the way modern FPS netcode does — authoritatively, and tolerant of lag.
The stack
The proof lives in a sim-rig built in eleven verified slices: SITL flight → fly-by-intent gateway with failsafes → live telemetry HUD over WebRTC → a real internet WAN loop through Cloudflare tunnels → a two-drone server-side tag game → both pilots over the WAN → match rules and mid-round rejoin → spectator feeds → a one-URL browser spectator page. Every slice ships with a headless PASS gate. Frontend is Next.js and Three.js; control plane is Python and MAVLink.
Where it's at
The business model, latency budget, legal path, and fleet economics are all researched and written up; the sim-rig proves the entire control path in software on a Raspberry Pi for $0 in hardware. The next milestone is Phase 0 — one netted bay, three drones, and a cross-country latency measurement that will either kill the idea or green-light a paid alpha.