Pre-departure.
The night-before preparation determines how the mission day flows. Forgotten gear becomes a wasted trip; insufficient battery capacity becomes a halted mission. Most mission-day failures are traceable to incomplete preparation the evening before. Cohort default pre-departure follows a structured sequence — equipment check, weather verification, route confirmation, communications setup.
The night-before checklist:
Morning-of departure:
- Verify weather one more time. Conditions can change overnight, especially in wet season.
- Pack the truck in this order: drone bag (cushioned, on top), batteries (in fireproof LiPo bag), field bag (tools), laptop bag, water and snacks.
- Confirm cooperative contact by phone/SMS. Verify they're still expecting you, the AOI is accessible, no surprises.
- Set departure timing to arrive at the site 60-90 minutes before planned takeoff. The buffer absorbs traffic, last-minute equipment issues, and site walkthrough.
The "everything is in the truck" check
Before leaving the driveway, take 30 seconds to verbally confirm each major item is in the truck:
- "Drone... yes."
- "Batteries — five packs... yes."
- "Radio... yes."
- "FPV goggles... yes."
- "Laptop and charger... yes."
- "Field bag... yes."
- "Phone with mission files... yes."
Sounds silly. Prevents the moment 90 minutes into the drive when you realise the radio is on the kitchen table. A later-cohort alumna lost a half-day mission this exact way; the verbal-check habit dates from then.
Arrival and site setup.
You arrive at the site. The cooperative manager is expecting you. The AOI looks like the satellite imagery suggested but with surprises. The first hour at the site is mostly social and logistical, not technical — meeting people, walking the boundary, choosing the launch zone, briefing observers. Skipping any of these creates problems during the mission you can't solve from the air.
The arrival sequence:
Cooperative leadership courtesies
The mission is happening on cooperative land at cooperative invitation. The relationship is the work's foundation. Cohort default courtesies that build long-term partnerships:
- Always greet leadership first, even if they're not actively involved in the mission. Acknowledge their role.
- Bring small contributions — coffee, biscuits, sometimes lunch for the team if a long mission. The cost is trivial; the gesture matters.
- Brief the leadership on what you're doing — even if they don't want technical detail. They want to know what to tell their members.
- Ask before flying over occupied areas, even if it's their land. Some members may be uncomfortable with overflights of their specific houses or fields.
- Document the visit — short post-mission summary sent to leadership within 24 hours. Builds trust for future work.
These aren't optional courtesy in cohort culture; they're operational requirements. Graduates who skip this layer find themselves not invited back to recurring sites.
Verifying mission file against site reality. The mission was planned at home using satellite imagery. Walking the AOI may reveal that adjustments are needed:
- Boundary differs from planned: the cooperative manager points out the actual boundary, which is 30m off from the satellite-image AOI. Re-plan the lawnmower lines to match the real boundary.
- New obstacle in the AOI: a building under construction, a new fence, a parked vehicle. Adjust the flight path to avoid; potentially reduce the AOI temporarily.
- Weather different from forecast: wind blowing perpendicular to planned line direction. Rotate the lawnmower pattern 90° (Section 4 of missions.html on lawnmower direction).
If the mission needs significant re-planning, do it at the site with the laptop. The 60-90 minute arrival buffer accommodates this. Don't fly a mission file that doesn't match site reality — the captured imagery won't cover what the client expects.
Pre-flight protocol.
The technical pre-arm checklist (fc-setup.html Section 10) runs before the first flight. This section adds the operational layer: confirming observers in position, communication standards, abort criteria, mission upload verification. The technical checklist verifies the drone; this section verifies the operation.
The pre-flight protocol runs in two parts:
The five operational pre-flight checks:
The five abort triggers worth memorising
Cohort default abort triggers, named explicitly so they become reflexive:
- Wind exceeds threshold for your build (~25 km/h sustained for 5" cohort default).
- Battery hits warning voltage (3.5V/cell) — RTH triggers automatically; pilot doesn't fight it.
- GPS satellite count drops below 6 in autonomous mode — switch to Angle (manual) immediately.
- Link quality degrades (RSSI below -100 dBm) — pause forward progress; if degrading further, RTH.
- Pilot loses visual contact with the drone — RTH or take manual control to bring it back into view.
Speaking these out loud at pre-flight commits to them. Speaking them out loud during a mission triggers the action. The discipline of verbalising is what makes abort decisions fast under pressure.
Once both pre-flight phases are complete, the mission is ready to launch. Total time from arrival to first takeoff: typically 60-75 minutes. Some of this overlaps with arrival activities (Section 2); after the first mission, between-flight turnaround (Section 5) is much faster.
In-mission operations.
The drone is in the air. The pilot watches it visually; the observers watch the surroundings; the configurator shows mission progress. The next 8-10 minutes are all about staying engaged when nothing seems to be happening — autonomous missions are operationally boring but punish lapses in attention. Communication discipline, role separation, and abort-trigger reflexes are the operational tools.
Role separation during a mission:
Communication standards during a mission
Cohort default communication patterns to keep the team coordinated without overwhelming the pilot:
- Pilot speaks first, always. Status reports every 30 seconds in calm conditions; more frequently if anything looks marginal.
- Concise format: "voltage X, line Y of Z, status Q". E.g., "voltage 3.78, line 5 of 8, normal".
- Observer interrupts only for hazards — not for general observations. "Person at north edge" is interrupting; "nice view of the cacao trees" is not.
- Decision-quality questions are escalated to the pilot: GS notices something marginal, asks "voltage at 3.62, are you continuing?"; pilot decides.
- Quiet is okay: silence after the pilot's status report means everyone's monitoring. Don't fill silence with chatter.
program graduates report that this discipline keeps mid-mission attention up across all team members. Without it, missions either go quiet (no one watching) or loud (everyone talking, pilot distracted).
When something needs immediate intervention:
- Take manual control. Switch out of mission mode to Position Hold (PosHold) or Angle. The drone stops following the planned mission and responds to stick input.
- Bring it home. If the situation is recoverable, fly back to the launch zone manually. If RTH is reliable, switch to RTH mode.
- Land safely. Calm landing in the original launch zone or the nearest emergency landing zone (identified during Section 2 setup).
- Pause and assess. Don't immediately re-launch. Discuss what happened, decide whether to continue the day, fix what's fixable, then proceed.
The "intervene then assess" sequence prevents a small problem from becoming a big one. Don't try to "ride out" a marginal situation hoping it resolves; bring the drone home, then think.
Battery turnaround.
Multi-battery missions need efficient turnaround between flights. Land, swap, verify, re-launch — the rhythm of the day. Long turnarounds eat the sun-angle window; sloppy turnarounds skip safety checks. Cohort default targets ~10 minutes between flights, with longer breaks after every 2-3 batteries to manage attention fatigue.
The turnaround sequence:
Mid-day decision points
After every 2-3 batteries, a longer break (~15-20 minutes) is worth taking even if the team feels fine. Three reasons:
- Pilot attention drift. Concentration on autonomous-mission monitoring degrades over the morning. A break re-baselines attention.
- Decision opportunity. Conditions change — wind picks up, clouds build, someone realises they need to leave early. The break is when these get noticed and addressed.
- Imagery spot-check. Pull the SD card, scroll through recent photos on the laptop. Verify capture is happening as expected. Catching imagery problems mid-day allows re-flying; catching them at end-of-day means another trip.
program graduates report the mid-day break is what separates "tired but completed" from "tired and made a mistake." Build it into the day's plan.
When to call it: situations that warrant ending the day even if planned missions remain:
- Wind exceeded threshold for 30+ minutes with no improvement in forecast — conditions likely to stay bad.
- Equipment failure that can't be field-repaired — broken FC, damaged motor, etc. Don't fly with marginal equipment.
- Pilot or team fatigue to the point of attention loss — calling missed waypoints, slower reactions, eye strain.
- Sun angle dropping out of the usable window — late-afternoon flying produces lower-quality NDVI; better to return another day.
- Safety concern not previously noted — bystanders refusing to clear, livestock unmanageable, conditions deteriorating.
Calling a day early is not a failure — pushing through bad conditions is. Cohort instructors specifically train this: the decision to stop is a sign of maturity, not weakness. The mission can be resumed another day; a wrecked drone or bad imagery cannot be undone.
End-of-day.
The last mission lands. The work isn't done. Imagery verification, equipment teardown, and the cooperative-manager debrief still need to happen on-site — leaving these for "back at home" is how missions end up needing repeat trips. The end-of-day phase is also when the day's lessons get captured for next time.
The end-of-day sequence:
The on-site imagery review
The 10-minute on-site imagery review is the difference between "mission complete" and "mission needed re-fly." What to look for:
- Completeness: scroll through chronologically. Is there a continuous capture sequence covering the entire AOI? Or are there gaps where the drone wasn't capturing for some reason?
- Exposure: are all photos similarly exposed? Sudden bright/dark frames suggest the camera adjusted incorrectly mid-flight; may stitch poorly.
- Sharpness: zoom into a few representative photos. Are they sharp? Or blurred from vibration / motion?
- Coverage: are the AOI corners and edges captured? Or did the mission cut short before reaching them?
If any of these fail, decide whether to re-fly while still on-site (preferable) or schedule a return trip (last resort). A later-cohort alumna once caught a partial-mission failure during this on-site review — re-flew the missed section with the spare battery while the cooperative manager prepared lunch. The 30-minute fix prevented a 2-hour return trip.
Post-mission home work happens after returning home, not at the site:
- Battery storage: discharge to storage voltage (3.85V/cell) before storing; mark each pack's cycle count.
- Imagery processing: see processing.html (proposed) for the WebODM stitching workflow and NDVI calculation.
- BlackBox archival: copy BlackBox files from SD card to organised folder structure on the laptop. Date- and mission-named.
- Client deliverable preparation: stitch the imagery, generate the NDVI overlay, prepare the deliverable per the agreement made during requirements gathering (missions.html Section 1).
- Field log review: re-read today's field log notes; transfer any lessons-learned to the persistent notes for future missions.
The mission day ends when the truck is parked at home and equipment is unloaded. The actual delivery to the client comes later — usually within 48-72 hours of the mission for cohort default work. The client agreement specifies the timing.