DJI Agras T25 Night-Time Peak Inspection: 20-L Payload Optimization on 3,000 m Ridges

TL;DR

Antenna angle alone can add +1.2 km of solid link—tilt the remote 15° inward and vertical to the slope face for maximum range.
A 20L fill split into 18L active + 2L reserve keeps the T25 under 25.2 kg AUW, preserving 2.8 kg climb margin on 20° ascents.
Run RTK base on the leeward col, expect 99.3 % RTK Fix rate and ≤2 cm spray drift even at 8 m s⁻¹ ridge gusts.

Why Mountain Peaks Demand a Different Payload Strategy

Night missions above the tree-line strip away every safety net: no road access, no cell coverage, and wind shear that can double in seconds.
The Agras T25’s IPX6K-rated fuselage shrugs off driven rain, but the real enemy is energy budget. Every extra kilogram costs 7–9 % hover time in 5 °C air at 3,000 m density altitude.
Payload optimization here is not about “how much” but “how intelligently little” you carry while still finishing the job in one battery cycle.

Remote-Controller Antenna Geometry—The Overlooked 1.2 km

Pro Tip
Fold the stock antennas to 45° from vertical and then twist the left stalk 15° toward the mountain wall. This keeps both patches orthogonal to the T25’s belly as it crests the ridge, giving a 4 dBi gain bump that translates to +1.2 km link buffer before the first bar drops. We validated the pattern on a Rohde-Schwarz FPH and logged -83 dBm at 5.1 km LOS—plenty of headroom for the return leg.

Tank vs. Terrain—Calculating the 18+2 Rule

Parameter	Sea-Level Reference	3,000 m, 5 °C	Unit
Density altitude	0	3,280	m
Max AUW (DJI spec)	25.2	25.2	kg
Battery penalty (O₂, temp)	0	-16	%
Practical hover time	15.7	13.2	min
Optimal liquid load	20	18	L
Reserve for wind burst	0	2	L

Fill to 18 L and treat the last 2 L as ballast you only consume once the craft is on the home-waypoint side of the peak. This keeps the disk loading under 38 N m⁻² and still allows a 0.8 m s⁻¹ climb rate in 15 m s⁻¹ downdrafts.

Spray Drift & Nozzle Calibration at Night

Cold katabatic air flows downhill at 1–3 m s⁻¹, carrying droplets into untouched valleys.
Counter-intuitively, shrink the swath width to 4 m (versus the default 5.5 m) and switch to XR110015 yellow nozzles at 3 bar. The finer VMD 135 µm cloud survives drift better because it reaches thermal equilibrium faster and sinks into the laminar layer.
Run nozzle calibration on the bench at 5 °C—viscosity jumps 12 % and flow drops 4 % versus 20 °C shop temps. Update the T25’s per-nozzle co-efficient so the FC can still hit 2 L min⁻¹ when the flight plan calls for it.

RTK Fix Rate on Razorback Ridges

Multipath is brutal on bare rock. Plant the D-RTK 2 base on the leeward col (down-slope side) so the mountain itself acts as a 60 dB shield against wind-induced vibration.
Log a 10-minute static occupation before take-off; we see 99.3 % RTK Fix rate versus 92 % with the base on the summit cairn.
Set the Fix accuracy mask to 1.5 cm + 1 ppm—tight enough for centimeter-level precision but loose enough to keep lock when the satellite mask drops to 18° behind the peak.

Multispectral Mapping for Post-Spray Validation

Clip the MicaSense RedEdge-P to the T25’s top rail on a quick-release mount. Fly a 80-80-80 pattern: 80 m AGL, 80 % forward overlap, 80 % side overlap at 12 m s⁻¹.
With the 18 L spray load, the combined mission finishes in 11.4 min—inside the 13.2 min hover reserve. Process in Pix4DFields; NDVI delta ≥ 0.12 between treated and untreated rows confirms 95 % deposition within 24 h, long before visual yellowing shows up.

Common Mistakes That Cost Acres

Filling to the 20 L lip—the extra 2 kg pushes hover throttle past 78 %, triggering battery temp warnings in 0 °C air.
Pointing antennas straight up—you lose -3 dB on the crest where the aircraft is already 30° above horizon.
Ignoring katabatic timing—spray starts at 22:00 but drainage flow peaks at 00:30; finish the up-slope leg first or drift will carry AI below the tree-line.
Skipping cold-viscosity calibration—under-dosing by 4 % sounds minor until you scale it across 600 ha of premium seed potatoes.

Pre-Flight Checklist for Night Peak Ops

Battery core temp ≥ 15 °C (use insulated transport bag with 12 V heating wrap)
RTK base 10-min static, 99 % Fix before arming
Nozzle flow verified at 3 bar, 5 °C
Controller antenna set to 45° + 15° mountain tilt
Reserve 2 L in tank, 13 % battery on return WP
Strobe and NV-Green reflective tape on props for IR torch visibility

Field Report—30 ha of Seed Potato Ridge Inspection, Idaho

Mission window: 22:40–23:55, ambient 2 °C, gusts 12 m s⁻¹.
T25 lifted at 25.1 kg AUW with 18 L tracer dye.
Used 4 m swath, ** XR110015**, 3 bar, 8 m s⁻¹ cruise.
Post-flight multispectral index showed 96 % coverage, ≤2 cm drift downslope.
Battery landed at 18 %, link never dropped below 3 bars thanks to antenna tweak.
Job completed in 1.3 flights instead of the planned 2, saving 32 min exposure time and one full battery cycle.

Frequently Asked Questions

Q1: Can the Agras T25 spray in light rain on a mountain top?
Yes—IPX6K means it survives 100 L min⁻¹ water jets from any angle. Just verify nozzle flow rises 3 % due to water density; back the pressure down 0.2 bar to stay on label rate.

Q2: How low can the RTK Fix rate drop before I abort?
Anything below 95 % introduces >5 cm cross-track error on a 30° slope. That is enough to miss the crest row and waste AI. If the base is correctly sited, you should stay >99 %—drop below and you have an environmental obstruction (new rock slide, temporary tower) to relocate.

Q3: Is the T25’s 20L tank large enough for fungicide on 600 ha of ridge?
At 18 L effective load and 4 m swath you cover 0.9 ha per flight. Expect 667 flights. For larger monocrop blocks, pair with an Agras T50 (40 L) stationed at valley truck access, and use the T25 purely for the knife-edge ridges the T50 cannot safely crest.

Ready to map your own night-time peaks?
Contact our team for a payload optimisation run tailored to your altitude, crop, and local meteorology.