Requirements first.
The client's actual need, expressed in their language, before any drone-side planning. Resolution requirements derive from this; altitude, overlap, battery count, and timing all derive from resolution. Skipping the requirements conversation produces missions that satisfy the drone but not the client — the most common mode of mission failure in alumni reports.
The conversation with the client (or partner cooperative, or your own farm if you're flying for yourself) covers four things:
The "what's normal" baseline question
Often missing from the explicit requirements but worth asking: "What does a healthy version of this field look like to you?" Many clients can't articulate disease symptoms in the abstract but can describe the contrast with healthy crops they've seen.
This is also where a client may reveal that they have prior survey data ("the agriculture office did this last year") which becomes a useful baseline. Two surveys six months apart show change; one survey alone shows only a snapshot. Ask whether earlier baseline data exists.
Documenting requirements. Write the requirements down before planning begins. The download kit at the top of this page includes an AOI checklist form trainees use during cohort training. The form's value is partly in the answers and partly in forcing the conversation that produces them.
Some requirements you cannot satisfy with cohort default builds. Be honest about that in the requirements conversation, not after a failed mission. A client wanting 0.5 cm/pixel resolution across a 50 ha cacao plantation is asking for something a cohort default 5" build cannot reasonably deliver. Either suggest scaling down the resolution, scaling down the area, or referring them to a partner with research-grade equipment.
Resolution & altitude.
The relationship between flight altitude and ground sample distance (GSD) is the single piece of math that drives most other planning decisions. Higher altitude = wider coverage per flight line = fewer batteries = faster mission, but worse resolution. Pick altitude from required GSD, not from convenience.
The GSD formula:
GSD (cm/pixel) = (sensor pitch in μm × altitude in m) / (focal length in mm × 10)
For the cohort default NDVI camera (Runcam Phoenix 2 with NDVI conversion, ~3.0 μm sensor pitch, 2.1 mm focal length):
The cohort default: 60-80 m altitude for most NDVI surveys with cohort default 5" build. This produces 8.6-11.4 cm/pixel resolution — sufficient for cooperative-scale crop health assessment, with manageable flight times for typical 5-20 ha AOIs.
Deviate from the default only with reason:
- Lower (30-50m) if the client needs individual-tree assessment or if the AOI is small enough that flight time isn't a constraint.
- Higher (90-120m) if the AOI is large (>30 ha) or time-constrained, and field-scale rather than tree-scale resolution suffices.
Required overlap interacts with altitude. Lower altitudes need more flight lines for the same coverage (because each line covers a narrower strip of ground). Section 3 covers the overlap math; the lines-per-hectare column above is calculated using cohort default 75% sidelap.
Coverage & overlap.
Aerial imagery for photogrammetric stitching needs overlap between adjacent images — both along the flight path (frontlap) and between adjacent flight lines (sidelap). The overlap is what lets the processing software identify common features and stitch images together. Insufficient overlap produces stitching gaps; excessive overlap wastes flight time. Cohort default is 75% frontlap and 65% sidelap.
The two overlaps:
- Frontlap (also called forward overlap or end overlap) — the percentage of consecutive images along a single flight line that overlap. The drone is flying forward; consecutive images of the ground share a percentage of pixels.
- Sidelap (also called side overlap) — the percentage of overlap between adjacent flight lines. Not consecutive images on one line; rather, the right edge of line N's images overlaps with the left edge of line N+1's images.
Both are expressed as percentages and both are commonly somewhere between 60-85%. The cohort defaults:
Why these specific values? 75/65 is in the sweet spot where cohort default builds reliably produce stitchable imagery with WebODM (the cohort default processing tool) without excessive flight time. Higher values (85/75) produce more reliable stitching but increase flight time by ~30%; lower values (65/55) cut flight time but increase the rate of stitching failures, which costs more time on re-flights than was saved.
The capture interval calculation — how often the camera should fire — derives from frontlap and flight speed:
Capture interval (s) = (footprint_length × (1 - frontlap)) / flight_speed
where footprint_length = (sensor_height_in_pixels × GSD_in_m) / 1
For the cohort default at 60m altitude (GSD 0.086 m/pixel), Runcam Phoenix 2 sensor (~720 vertical pixels), and 75% frontlap:
- Footprint length = 720 × 0.086 = 61.9 m
- At 5 m/s flight speed: capture interval = (61.9 × 0.25) / 5 = ~3.1 seconds
- At 8 m/s flight speed: capture interval = (61.9 × 0.25) / 8 = ~1.9 seconds
The mission planner generates these values automatically; you don't need to calculate manually. But understanding the relationship helps when you need to debug a mission with stitching gaps — they often come down to one of these inputs being miscalibrated.
Sidelap and flight line spacing. The lines-per-hectare numbers in Section 2 assume cohort default 65% sidelap. If you change sidelap, recalculate. The formula:
Line spacing (m) = footprint_width × (1 - sidelap)
where footprint_width = (sensor_width_in_pixels × GSD_in_m) / 1
For 60m altitude, Runcam sensor (~1280 horizontal pixels), 65% sidelap:
- Footprint width = 1280 × 0.086 = 110 m
- Line spacing = 110 × 0.35 = 38.5 m
- Lines per ha = 100m / 38.5m ≈ 2.6 lines (round up to 3 for safety)
The cohort coverage-math spreadsheet (in the download kit) has these calculations built in for the cohort default cameras and several alternatives. Plug in your altitude, AOI dimensions, and target overlap; it produces line count, line spacing, and total mission distance.
Flight patterns.
Four flight patterns are commonly used for survey work. Lawnmower (parallel back-and-forth lines) is the cohort default and works for ~85% of missions. The alternatives — perimeter-first, spiral, terrain-following — solve specific problems where the lawnmower doesn't work well. Pick the simplest pattern that gets the coverage you need.
The lawnmower direction matters
For most AOIs, lawnmower lines should run perpendicular to the dominant wind direction. Reasons:
- The drone's ground speed varies more on lines parallel to wind (faster downwind, slower upwind). This creates inconsistent capture intervals and patchy frontlap.
- Crosswind flight has more consistent ground speed. Capture intervals stay steady.
- Heading correction (yaw to maintain track) is easier on perpendicular lines than parallel ones.
Check the wind direction at the start of mission planning. If conditions change, adjust the line direction. Cohort default mission planners (in the download kit) include wind-direction inputs that auto-rotate the lawnmower pattern.
Pattern selection by AOI shape:
- Square or roughly rectangular (most cooperative plots) — lawnmower, lines perpendicular to wind.
- Long thin (river edge, road corridor, irrigation channel) — perimeter-first, then narrow lawnmower if interior detail needed.
- Circular (irrigation pivot, pond perimeter) — spiral OR lawnmower over a square that contains the circle (lawnmower simpler).
- Irregular polygon — lawnmower over the bounding rectangle. Trim the resulting imagery to the actual AOI in post-processing. Simpler than trying to fly the irregular shape exactly.
- Mountainous or significantly sloped — terrain-following lawnmower if elevation data is reliable; otherwise lawnmower with conservative altitude (above the highest point + safety margin).
For cohort default missions, expect to use lawnmower 85-90% of the time. The other patterns are tools for specific situations, not first choices.
Battery budgeting.
Cohort default 5" build with a 1500 mAh 4S pack flies for ~10 minutes of usable mission time. Each mission needs to fit within that envelope or be split across multiple batteries. Underestimating battery time is the most common reason cohort missions abort mid-flight — get this right at planning time, not in the air.
The mission time budget breaks down into:
- Pre-mission (after takeoff, before first capture line): ~30 seconds for climb, GPS lock confirmation, transit to first waypoint.
- Capture phase: total flight distance / flight speed. This is the bulk of the mission.
- Return-to-home: the drone needs to fly from the end of the mission back to landing. ~30-60 seconds depending on AOI distance from launch point.
- Safety margin: 20% reserve, hard rule. Never plan to land at empty.
For cohort default 5" build with cohort default 1500 mAh 4S pack:
Translation to flight distance. At cohort default 5 m/s flight speed, ~7-9 minutes of capture phase = 2,100-2,700 meters of flight distance. Working backward from the mission's total flight distance:
Battery count needed = ceiling( total_mission_distance_m / 2400 )
For 20 ha AOI at 60m altitude, 65% sidelap, 5 m/s = ~3,200 m total mission
→ 3200 / 2400 = 1.33 → 2 batteries needed
Practical battery counts for typical cohort missions:
- 1-5 ha plot: 1 battery, sometimes 2 if very high resolution required
- 10-15 ha cooperative plot: 2 batteries
- 20-30 ha cooperative survey: 3 batteries
- 40-50 ha large cooperative: 4-5 batteries; consider splitting across two flying days
- >50 ha: definitely multi-day; reconsider whether higher altitude can reduce flight time
The cohort recommendation: always carry one more battery than your plan calls for. A 3-battery mission becomes a 4-battery flight day. The extra pack covers test flights at the start, equipment problems, and the occasional re-fly of a section that didn't capture cleanly. Adds ~₱2,500 of pack on the truck for the day; saves a return trip to the site.
Battery age affects budget
Cohort default 1500 mAh packs after ~50 cycles deliver closer to 1300 mAh of usable capacity. After ~100 cycles, ~1100 mAh. The mission planning above assumes fresh packs; aged packs need a reduction in mission length proportional to capacity loss.
Practical rule: track each pack's cycle count in the fleet logbook. When a pack drops below 80% of new capacity (typically ~70 cycles), retire it from mission work and use only for training/practice. Don't mix new and aged packs in the same mission — it complicates the math and creates risk.
Timing.
Sun angle, wind windows, crop growth stage, and seasonal patterns all affect mission outcomes. NDVI imagery in particular is sensitive to sun angle — flying at the wrong time of day produces shadows that wreck the analysis. For cohort default Mindanao operations, the prime survey window is roughly 10 AM to 2 PM, with weather and seasonal modifications.
Sun angle for NDVI:
- Below 30° from horizontal (early morning, late afternoon): long shadows, low light, NDVI values compressed by shadow noise. Avoid for serious analysis work.
- 30°-50° (mid-morning, mid-afternoon): workable but shadows still significant. Acceptable for ground-truth-supplemented analysis.
- 50°-80° (late morning to early afternoon): cohort default range. Shadows minimised; uniform illumination across canopy; NDVI values most reliable.
- Near 90° (solar noon, equatorial regions): excellent illumination but sometimes too much glare on shiny canopies (banana, coconut). Brief mid-day pause possible if specular reflection dominates imagery.
For Mindanao (latitude ~7° N), solar noon happens around 11:45-12:15 depending on season. Sun angle exceeds 50° from approximately 9:30 AM to 2:30 PM year-round. The cohort default flight window is 10:00 AM to 2:00 PM — comfortably inside the high-sun band, with a 30-minute buffer on each end for setup and teardown.
Wind windows. The cohort default 5" build flies acceptably up to ~25 km/h sustained wind; struggles above that; doesn't fly safely above ~35 km/h. The Mindanao wind pattern in agricultural regions:
- Morning (6-9 AM): typically calm; light land-breeze in coastal regions.
- Mid-day to afternoon (10 AM-3 PM): wind increases as land heats; sea breezes develop in coastal regions; thermal updrafts in inland regions. Peak typically 12-2 PM.
- Late afternoon to evening (4-6 PM): wind moderates as temperature drops.
This creates tension with the prime survey window (10 AM-2 PM = peak wind). The compromise is usually: fly during the high-sun window, accept that wind handling will be needed, and abort if wind exceeds the build's capability rather than push through.
Seasonal patterns:
- Dry season (December-May): most reliable flying weather. Steady winds, predictable patterns, low rainfall risk. Best for major missions.
- Wet season (June-November): afternoon thunderstorms common; mornings often cloudy; wind patterns less predictable. Plan missions for early in the window (10-11 AM), be ready to abort if weather develops.
- Typhoon season (July-October peak): occasional total no-fly periods of 2-7 days as systems pass through. Build mission schedules with 2-week flexibility windows during this period.
Crop growth stage matters too
NDVI values vary across the crop's growth cycle. Surveys at different stages show different things:
- Early growth (post-planting, pre-canopy-closure): bare soil dominates, NDVI low and patchy. Useful for stand establishment assessment.
- Mid-growth (canopy closing, vigorous growth): NDVI rising rapidly. Best for irrigation pattern analysis.
- Peak growth (full canopy, pre-flowering): NDVI peaks, most spatial variation visible. Best for stress detection and disease scouting.
- Senescence (post-flowering, maturation): NDVI declining as crops mature. Useful for harvest timing.
Discuss growth stage with the client during requirements gathering (Section 1). A survey timed wrong for the question being asked produces ambiguous data.
Safety planning.
Mission planning isn't complete until the safety plan is documented. Line-of-sight requirements, emergency landing zones, observer placement, and proximity considerations are all easier to think through at planning time than to figure out in the air. The cohort safety planning template (in the download kit) walks through the standard checklist.
Visual line of sight (VLOS). Philippine CAAP regulations require the pilot to maintain visual line of sight with the drone during commercial operations (without specific Beyond-VLOS authorisation). Practical implications for mission planning:
- Maximum useful distance from pilot: ~500 m for cohort default builds. At 500 m, the drone is a small dot but still visible. Beyond this, visual contact gets unreliable.
- If AOI extends beyond 500 m from a single pilot position: split the mission into multiple flights from different launch points within the AOI, OR use a mobile observer (Section 7's observer placement notes).
- If visual contact is lost during a mission: this is failsafe territory. RTH is the fallback. Re-establishing contact via observer is the recovery path. See safety.html for full procedure.
Emergency landing zones. Identify before the mission. Two questions:
- If something goes wrong mid-mission (low battery warning, link loss, motor issue), where can the drone land safely?
- Are these zones inside the AOI, on the way to/from RTH, or outside the mission corridor entirely?
For typical cooperative surveys: cleared areas around the launch point, harvested fields, dirt roads, and open grassland all work. Avoid: water, dense canopy (drone gets stuck), populated areas, structures, livestock pens. Mark landing zones on the AOI map. Cohort training includes practising landings to specific marked zones, not just back to the takeoff point.
Observer placement. For larger missions, an observer in addition to the pilot adds significant safety:
Proximity considerations. The mission shouldn't bring the drone unsafely close to:
- People — minimum 30m horizontal separation from non-participants (CAAP requirement). For cohort default lawnmower patterns, plan AOI boundaries so flight lines don't pass over occupied areas.
- Structures — minimum 30m horizontal separation from buildings, towers, telephone wires. Telephone wires in particular are easy to miss visually until the drone is too close.
- Livestock and pets — drone noise spooks animals; plan flight paths to avoid sustained overhead presence over enclosures.
- Other drones or aircraft — if any other drone operations are happening in the AOI vicinity, coordinate. Two drones in the same area without coordination is a collision risk.
- Restricted airspace — verify the AOI isn't inside or near no-fly zones (airports, military installations, prisons, certain government facilities). The CAAP NoTAM map covers most of these; safety.html has full reference.
The cohort safety planning template includes a checklist that covers all of these systematically. Filling it out takes ~10 minutes; doing it badly causes hours of avoidable problems.
Mission file generation.
With requirements, altitude, overlap, pattern, battery budget, timing, and safety planning complete, the final step is generating the actual mission file the FC will execute. This is the most mechanical step in the planning process — most of the decisions have been made; the file is just a structured record of them. INAV and ArduPilot use different formats; both are workflow-compatible with cohort default tools.
The cohort workflow uses two tools depending on FC firmware:
- For INAV (cohort default 5"/7" builds): INAV Configurator's mission planning tab, or
UgCSfor more sophisticated planning. INAV mission files use the.missionformat. - For ArduPilot (cohort default 10" builds): Mission Planner standalone application, or
QGroundControlfor cross-platform work. Format is.waypoints.
The basic mission file workflow (INAV):
- Open INAV Configurator, connect to the FC via USB or Bluetooth.
- Navigate to the Mission Control tab.
- Drop a starting waypoint at the planned launch point. Type:
WAYPOINT. Altitude: cohort default mission altitude (60-80m). - For lawnmower patterns: use the planning tool's "Survey" feature. Define the AOI polygon by clicking corners on the map; specify altitude, sidelap, and flight speed. The tool auto-generates the lawnmower lines.
- Add a final RTH waypoint at the end of the mission. Type:
RTH. - Verify the mission visually: line spacing looks right, AOI is fully covered, no waypoints in invalid positions (over water, structures, etc.).
- Save the mission file to disk (for backup) and upload to the FC.
What to verify before flight:
{client}-{date}-{altitude}.mission.Offline mission planning
Many cohort flying locations don't have reliable cellular data. The mission planner tools handle this in different ways:
- INAV Configurator: caches map tiles. Plan the mission with map data while online (at home, before traveling); the cached tiles work offline at the site. Cache covers ~30km radius around any location you've previously loaded.
- Mission Planner (ArduPilot): similar map caching; specifically download the AOI tiles before traveling.
- UgCS: requires online for initial planning but generates self-contained mission files that don't need the planner during flight.
The cohort default workflow: plan missions at home or at the cohort office before traveling to the field. The trip to the field is for executing missions, not planning them. Don't plan complex missions on a phone in a remote barangay — too easy to miss inputs without the bigger screen and reliable network.
Saving missions for re-use. Recurring surveys (monthly NDVI of the same cooperative, quarterly yield estimation) re-use the same AOI. Save the mission file with version control: {client}-{date}-{altitude}.mission. Next survey, load the existing file, update the date, verify nothing about the AOI has changed (new structures, new vegetation), upload.
Cohort engineering maintains a shared mission template repository for partner cooperatives that have signed on for recurring surveys. Alumni working with these partners pull templates from the shared repo rather than re-planning from scratch each time.