Farms Atlas
Our proprietary AI engine. The physics of your operation.
Developed by Farms, Atlas combines AI, mathematics and physical models to plan agricultural applications and evaluate their demands on time, equipment and resources.
Plan the core of your operation before mobilizing equipment and teams: how to apply, maneuver, return and support each work cycle.
- Flight dynamics
- Application and resources
- AI planning
Explore a recorded example.
Loading the operation's power grid.
Area and airstrip from the original file
Read the operation on the map
Choose the equipment
Explore the recorded Air Tractor AT-502B and DJI Agras T100 scenarios.
Distinguish transit from application
Follow passes, maneuvers and returns. The application state indicates liquid release.
Explore the complete cycle
Pause and scrub through time. For the drone, follow tank, battery and ground service.
01 / Farms Atlas
AI that plans. Simulation that explains.
Farms Atlas explores execution alternatives, evaluates their consequences using physical models and selects plans within the mission’s conditions.
Understand where time goes, which resources sustain the work and what changes between alternatives.
Takes in the scenario
Field, equipment, base or runway, dose, swath and support resources within the model’s scope.
Explores and evaluates alternatives
AI explores routes and work cycles within physical constraints and available resources.
Produces a plan you can inspect
Route, time, flights and resources, as supported by the model. Replay connects the numbers to operational stages.
02 / The science behind the engine
Applied mathematics. Physics in action.
Planning intelligence becomes useful when each alternative meets physics. Our engine evaluates how turns, changing payload, application and returns to base affect mission behavior and duration.
Motion with consequences
Mass, speed, acceleration and maneuverability are part of the problem. Aircraft dynamics account for aerodynamic forces, propulsion and wind. For drones, force and power requirements respond to motion and payload within the model's supported range.
Application along the route
Prescribed rate, effective swath, flow and activation connect movement to liquid release. Models with spatial analysis also let us investigate gaps, overlaps and off-target application.
Resources across the full cycle
Released liquid reduces the onboard supply; flight takes energy and time. Returns, refilling, battery swaps and waiting form part of the operation in models that support them. Ground time contributes to the result too.
03 / Equipment capabilities
Aircraft and drones. Every operation has its own context.
Equipment-specific models connect planning to each aircraft’s or drone’s capabilities and limits.
Agricultural aircraft
Investigate the relationship between runway, treatment area and trajectories. The cycle can include takeoff, transit, alignment, application passes, turns, return, landing and refilling.
- Flight dynamics, wind and changing payload.
- Time spent applying, maneuvering and in transit.
- Mission continuity between loads, where supported.
Drone applications
Investigate a fixed-base operation in which application, return and ground service form successive cycles. The tank, batteries and available support influence the pace of work.
- Motion, payload and flight energy within the model's supported range.
- Returns, refilling and battery swaps.
- Battery availability, charging and ground waiting.
The application plan
The core of your operation. Planned in detail.
An application plan connects passes, maneuvers and returns to equipment constraints and available resources. Examine those choices before mobilizing the operation.
Switch between aircraft and drone and reveal the sequence with the slider. In the replay, connect the route to the time and resources the plan requires.
- Application in green
- Turns and transit in yellow
Air Tractor AT-502B
04 / Atlas in action
From planning to replay. One operation, in detail.
This is the DJI Agras T100 recording in the replay above: an irregular field, a fixed base, a prescribed dose of 10 L/ha and 3 batteries in the simulated scenario.
- Flights in the operation
- 13
- Total duration · min:s
- 155:54
- Released liquid · L
- 317.37
- Flight energy · kWh
- 19.33
Values from the simulated record, not a field trial. Duration includes service and support shutdown; energy refers to flight, not total ground-charging consumption. Liters released are not equivalent to liters deposited on target.
- 10:25
The first return
The first flight ends with 0.27 L in the tank and 26.9% battery. These values show its state on arrival.
- 12:06
The next flight begins
About 101 seconds separate the return from the next departure: ground-service time that contributes to the total.
- 155:54
The operation result
The result combines all 13 flights and ground support through shutdown.
Select DJI Agras T100. Timestamps use min:s, with fractional seconds omitted.
This replay shows flight, resources and events. Spatial dose analysis is not included.
05 / Comparisons that inform decisions
Which choice makes sense for this operation?
Compare alternatives within the same context, including ground support. These questions help guide the analysis.
| Question | Context to keep explicit | What to examine |
|---|---|---|
| How does field shape affect the work? | Geometry, swath, dose and equipment within the supported scope. | Passes, turns, transit and total duration. |
| Which equipment fits this scenario? | The same field and application objective, with each model's own parameters and support. | Duration, flight cycles, resources and limitations; nominal capacity alone cannot settle the comparison. |
| Where does ground service have the greatest impact? | The same mission, with base location and support resources stated. | Returns, service, battery waiting and time between flights. |
06 / Scientific rigor
Rigor behind every result.
Our planning AI is developed alongside mathematical and physical models, with numerical verification and well-defined comparisons.
Physical and numerical consistency
We check movement, state continuity and resource balances. Results are model estimates; validating field performance, deposition, drift or agronomic efficacy requires dedicated measurements.
Preserved context
Each result preserves its conditions, equipment and model version. Earlier recordings retain their original context.
Well-defined comparisons
Studies state assumptions, variables and metrics. These recordings use different scenarios; they establish neither equipment superiority nor a global optimum.
Articles explore scenarios, results and evidence through simulation studies.
Bring a question from the field. Let's build the scenario.
Talk to the Farms team about the field, equipment and decision you want to investigate. Together, we can assess the simulation scope that fits your case.