The autonomy that flies a live wildfire patrol in Maestro Fire runs in your browser first. Before a single drone leaves the ground, you draw your actual high-risk zone, set your fleet, and watch the whole patrol play out, coverage, relay timing, battery endurance, and detection alerts, against the real topology you'll fly. Same flight logic, same decisions, fed by a physics model instead of a live link. It's the difference between hoping a configuration holds and seeing it hold before it matters.
The same autonomy, in a browser
A live wildfire patrol asks a lot of a configuration. The fleet has to cover a defined zone without gaps, hand off cleanly from a drone running low on battery to one with a full charge, and keep a continuous watch over the ground without anyone on the ridge babysitting it. Get the fleet size, the swap timing, or the zone shape wrong and you find out mid-flight, the worst possible moment.
Maestro Fire lets you find out on the ground instead. The browser dry-run runs the exact autonomy that flies the live mission: the same coordination engine, the same battery-aware handoff logic, the same safety checks that govern conservative behaviour in wind. The only thing that changes is where the telemetry comes from. Instead of a live radio link, position and battery state come from a physics model. Everything downstream behaves identically, so what you watch in the browser is what the fleet will actually do.
That fidelity is the point. A dry-run isn't a cartoon of a patrol; it's the patrol, rehearsed. If a wind threshold triggers a conservative re-plan in the browser, the same logic triggers it in the field. If the fleet leaves a corner of your zone uncovered on screen, it will leave that corner uncovered in the air. You see the configuration's real behaviour while it's still free to change.
Validate against your actual zone
Demo terrain doesn't tell you whether your fleet can hold your hillside. So the dry-run works on yours. Draw the high-risk zone you're actually responsible for, the wildland-urban interface above a town, a forested ridge, a stretch of dry scrub, and Maestro Fire plans the patrol against that exact shape and size.
Three things become visible the moment the patrol starts moving:
- Coverage. Does the fleet sweep the whole zone, or does the geometry leave a blind spot? An awkward concave boundary or an over-large area shows its gaps in motion long before it shows them in smoke.
- Relay timing. Watch the handoffs. A standby drone climbs and transits to take over as the active drone's battery drains, the two never sharing the same airspace. If your fleet is too small for the area, you'll see the watch lapse, coverage drops while everyone's on the pad charging, and you'll see it as a gap in the timeline, not a surprise on shift.
- Battery endurance. Real airframes have real flight times, and the dry-run uses them. If your zone is too far from base, or too large for the endurance you've got, the rehearsal surfaces it as drones that can't make the round trip with margin to spare.
Maestro Fire recommends a sensible fleet layout for the zone and endurance you've entered, then you watch whether the recommendation holds. Change the fleet size, change the battery, redraw the boundary, and run it again. Each pass takes a few minutes, not a shift, so you can settle a configuration in an afternoon instead of learning it the hard way over a season.
One rehearsal, the pace you need
A speed control sets how fast the rehearsal runs relative to real time, and it's the same patrol underneath at every setting:
- Real time. One-to-one with the drone. Configure tomorrow's patrol, start it, and let it run alongside your pre-flight prep while relay cycles trigger at the true cadence. This is the closest thing to standing on the ridge watching the fleet, without the fleet being airborne.
- Fast-forward. See three or four full relay cycles in a few minutes when the pattern matters more than the pacing.
- Mid-length preview. A whole shift compressed to ten minutes or so, sized for a tabletop review with the team around a laptop.
- Compressed showcase. A full day of patrol in roughly a quarter of an hour, the default, ideal for confirming the pattern holds across many cycles.
- Fast scan. A multi-day patrol skimmed in minutes when you just want to confirm the watch never lapses.
Detection alerts, handoff timing, battery drain, and the fleet's response to wind all scale together, so a fast pass and a slow pass tell the same story at different resolutions. Run it fast to check the shape of the plan; run it at real time to feel its rhythm.
Detection, rehearsed too
Maestro Fire is built for early detection: spotting an ignition in a defined high-risk zone, under your own authorizations, while it's still small enough to act on, and confirming it with a second drone before anyone commits resources. The dry-run rehearses that loop. As the patrol sweeps the zone, simulated heat sources appear within sensor range and the detection-and-alert path fires exactly as it would in the field, the alert surfaces, the map pans to the location, the notification you've configured goes out.
You're not testing whether the model can see fire, you bring your own sensors and models for that. You're rehearsing the operational loop around the detection: how fast the alert reaches you, what your team does with it, whether a confirming drone gets eyes on the spot. Practising that loop on the ground means that when a real alert lands, the response is already muscle memory.
Try it
No install, no login. The demo opens with a small fleet patrolling at the compressed default, so a full-day scenario plays out in a few minutes. A few things worth doing:
- Switch to real time and watch a single relay handoff at true pacing, the clearest way to understand why the fleet hands off when it does and never leaves the zone unwatched.
- Draw your own high-risk zone over terrain you actually cover. Set the fleet size, pick an airframe, and run it, the rehearsal anchors on your shape instead of the default scene.
- Shrink the fleet by one drone and run it again. If the watch lapses, you'll see exactly where, the case for the larger fleet, made visible.
- Set a fixed patrol window and run it fast, a complete rehearsal, start to finish, including the relay cycles, in a couple of minutes.
Every recommendation the autonomy makes during the run is logged, the same audit trail a live mission produces. What you rehearse is what you fly.
The principle
There's no separate "demo logic" anywhere in Maestro Fire. The browser rehearsal and the live patrol share one autonomy stack, so the rehearsal is a faithful preview rather than an approximation. That's what makes it worth your time: you validate coverage, relay timing, and battery endurance against your real zone, settle on a configuration you trust, and only then put drones in the air, with the plan already validated on the ground.
Rehearse before you fly. Run a patrol on your own zone.