Two aircraft do not automatically double output. They may share the same bottleneck.
Distinguish formation from coordination
This is a design proposal, not flight-control software. Partitioning work among aircraft differs from cooperative control using mutual state. Confirm local conditions for concurrent operation, staffing, observation and communications before choosing an architecture.
Reserve more than treatment areas
Reserve time windows for field cells, transit corridors, takeoff areas, contingency landing sites and service facilities. Consider downwind effects between adjacent jobs. A reassignment must transfer completed and pending treatment areas to avoid double application.
Treat stale telemetry explicitly
Attach timestamps and expiry to each aircraft state. Do not release its reserved space immediately after link loss. Test contingencies that avoid simultaneous returns to one landing point, with pilot intervention appropriate to the aircraft and jurisdiction.
Queues determine scale
Evaluate maximum waiting time, charging queues, operator workload, rework and missing telemetry alongside coverage. Progress through simulation, controlled tests and limited fields, revisiting operating conditions and staffing with each increase in concurrency.
Separate multi-drone sensing from spraying capability
Olds College describes planned simultaneous drone flights and ground validation for rock mapping. That page does not establish a completed spraying swarm, profitable scaling or cross-vendor coordination. Project objectives must remain distinct from verified outcomes.
Sending an assignment does not confirm it
The following is a management design example, not aircraft-control instructions. An assignment record should include mission ID, plan version, field cell, aircraft, responsible person and start conditions. Separate a proposed assignment from field acceptance, avoiding duplicate assignment when acknowledgment is absent. Before reassignment, confirm the former operator’s state and reconcile completed work, then approve a new version. Use versions and sequence numbers to prevent old plans reviving after reconnection. Link field reservations to launch facilities, service staff and nearby workers’ schedules. An apparently vacant cell on a screen is insufficient evidence that work can begin.
Test inconsistencies after recovery
In a test environment, inject delayed positions, duplicate completion notices, clock discrepancies and operator handovers as well as link loss. Verify recovery from a hold, rather than merely displaying a red warning. Record who checked the actual field state and what evidence supported renewed authorization. Aircraft link-loss behavior remains governed by its approved procedures, not this management example. Also test simultaneous alerts to one person and whether required checks accumulate beyond their capacity. Treat restored connectivity and authorization to resume as different states.
| Test event | Management-record check |
|---|---|
| Old position arrives late | Do not roll back the newer state |
| Duplicate completion | Do not count the area twice |
| Operator handover | Transfer holds and unfinished coverage |
| Connectivity restored | Do not clear holds without field confirmation |
Evaluate accepted work rather than accumulated flight area
If two aircraft photograph the same cell, adding their covered areas does not demonstrate doubled productivity. We propose measuring unique accepted coverage, gaps, repeat capture and human review time. For application work, overlap can also be a treatment-quality problem and requires different accounting. Compare single- and multi-aircraft work under comparable field conditions and tasks, recording differences. A larger fleet is not ready for acceptance if charging queues or unreviewed alerts grow. Include completion within the agreed deadline and the ability to reconstruct the work history, alongside daily output.