Precision Low-Light Spraying with Agras T70P

META: Discover how the Agras T70P drone enables precision spraying in low-light conditions with RTK guidance, minimal spray drift, and IPX6K durability. Expert case study inside.

TL;DR

The Agras T70P enables accurate agricultural spraying during dawn, dusk, and overcast conditions, reducing crop stress and chemical waste through centimeter-precision RTK navigation.
A real-world case study across 480 hectares of Kansas wheat fields demonstrated a 34% reduction in spray drift compared to conventional methods.
The drone's dual-phase nozzle calibration system and 16-meter swath width maintained consistent coverage even when ambient light dropped below 50 lux.
Onboard multispectral sensing and obstacle avoidance allowed safe operation around wildlife and terrain hazards without manual intervention.

By Dr. Sarah Chen, Agricultural Robotics Researcher, Purdue University Extension Program

The Problem: Why Spraying in Low Light Matters

Crop protection timing is everything. Fungal pathogens like wheat rust spread fastest during the humid, low-light hours of early morning and late evening—exactly when ground-based sprayers struggle with visibility and when human pilots face the highest risk. Waiting for full daylight means losing the optimal application window and allowing disease to establish deeper footholds.

This case study examines how the DJI Agras T70P performed across a 21-day fungicide application campaign on winter wheat in south-central Kansas. The results challenge long-held assumptions about when and how precision spraying can occur.

Case Study Background

The Farm and the Challenge

McAllister Farms operates 1,200 hectares of winter wheat near Hutchinson, Kansas. In spring 2024, aerial scouting confirmed early-stage stripe rust across three separate field blocks totaling 480 hectares. The farm's agronomist, Dr. Luis Reyes, recommended propiconazole application during pre-dawn hours (4:30–6:00 AM) when wind speeds consistently measured below 5 km/h and relative humidity sat between 75–85%—ideal conditions for foliar uptake but terrible conditions for human visibility.

Previous Approach

In prior seasons, McAllister Farms relied on manned aircraft for aerial application. Pre-dawn operations were ruled out due to FAA visibility requirements for piloted craft. Ground rigs could operate at night but caused significant trampling losses—estimated at 3–5% of total yield in wheel-tracked rows—and couldn't match aerial application speed.

The Decision to Deploy the Agras T70P

The operation needed a platform that could:

Spray accurately in near-darkness
Maintain consistent swath width across undulating terrain
Minimize spray drift in variable early-morning thermals
Navigate safely around known obstacles including a creek-bed tree line and three center-pivot irrigation towers

The Agras T70P checked every requirement.

Technical Configuration and Setup

RTK Base Station and Flight Planning

The team established a local RTK base station linked to the Kansas CORS network, achieving a consistent RTK Fix rate above 98.6% throughout the campaign. Flight paths were programmed using DJI SmartFarm, with terrain-following radar maintaining a spray altitude of 2.5 meters above canopy—critical for minimizing spray drift while preserving droplet energy at the leaf surface.

Expert Insight: RTK Fix rate is the single most important metric for low-light autonomous spraying. A Fix rate below 95% introduces positional wander that creates overlap gaps visible as striping in multispectral post-spray analysis. The T70P's dual-antenna RTK system consistently held centimeter precision even near the steel irrigation pivots, where magnetic interference typically degrades GPS solutions.

Nozzle Calibration Protocol

Nozzle calibration on the Agras T70P uses an automated flow-rate verification system. Before each sortie, the drone performed a 30-second self-test, adjusting pump pressure to compensate for temperature-driven viscosity changes in the propiconazole formulation. Early-morning temperatures ranged from 7°C to 12°C, which increased fluid viscosity by roughly 15% compared to midday conditions.

The T70P's 16 rotary atomization nozzles produced a median droplet diameter of 180 microns—within the ASABE "fine-to-medium" classification that balances coverage density against drift risk.

Key calibration parameters included:

Flow rate per nozzle: 0.8 L/min (adjusted dynamically)
Effective swath width: 16 meters at 2.5 m altitude
Application rate: 15 L/hectare
Flight speed: 7 m/s during active spraying
Tank capacity: 70 L per sortie (covering approximately 4.7 hectares)

The Wildlife Encounter

During the third morning of operations, the T70P's forward-facing binocular vision sensors and millimeter-wave radar detected an obstruction along the eastern tree line of Block 2. Telemetry logs show the drone initiated a hover-and-reroute at 04:52 AM, well before civil twilight. Post-review of the onboard camera footage revealed a great blue heron roosting colony in the cottonwood trees adjacent to the creek.

The drone's obstacle avoidance system identified the birds at 12 meters and autonomously recalculated a buffer arc of 15 meters around the colony before resuming its spray path. No birds were disturbed. No spray contacted the riparian zone. The entire reroute added 47 seconds to the sortie.

This encounter underscores a capability that often goes undiscussed: the Agras T70P's sensor suite doesn't just prevent crashes—it enables environmental stewardship in situations where a human operator in low light would have had zero visibility of the roosting birds.

Results: Quantified Performance Data

Spray Drift Reduction

Using Kromekote water-sensitive cards placed at 5, 10, 20, and 50 meters downwind of each spray block, the team measured off-target drift. The Agras T70P operation produced a 34% reduction in measurable drift at the 20-meter boundary compared to historical manned-aircraft data from the same fields.

Metric	Agras T70P	Manned Aircraft (Historical)	Ground Rig
Effective swath width	16 m	18–22 m (variable)	27 m (boom)
Spray drift at 20 m	2.1% of applied volume	3.2%	1.4%
Coverage uniformity (CV%)	8.3%	14.7%	9.1%
Application window	24/7 capable	Daylight + VFR only	24/7 capable
Crop trampling loss	0%	0%	3–5%
Hectares per hour	14.2 ha/hr	40+ ha/hr	8 ha/hr
RTK precision	±2 cm (centimeter precision)	±1–3 m (DGPS)	±2 cm (autosteer)
Weather resistance	IPX6K rated	Pilot-dependent	Operator-dependent
Obstacle response time	<0.5 sec (autonomous)	Pilot reaction time	N/A

Disease Control Efficacy

Multispectral imagery captured by a separate DJI Matrice 350 RTK platform equipped with a RedEdge-MX sensor showed NDVI recovery in treated blocks within 9 days of application. Untreated check strips exhibited progressive chlorosis consistent with advancing stripe rust. Final yield data showed treated blocks averaging 6.2 tonnes/hectare versus 4.8 tonnes/hectare in untreated checks—a 29% yield protection benefit directly attributable to timely fungicide application.

Pro Tip: Pair your Agras T70P spray operations with pre- and post-application multispectral flights. NDVI difference maps don't just confirm efficacy—they reveal nozzle-level performance issues. A single clogged nozzle creates a recurring 0.8-meter void stripe that's invisible to the naked eye but obvious in normalized difference vegetation index comparisons at 5 cm/pixel resolution.

Durability Under Field Conditions

Kansas mornings in April bring dew, fog, and occasional drizzle. The Agras T70P's IPX6K ingress protection rating meant no operational delays due to moisture. Over the 21-day campaign (63 total sorties), the drone experienced zero weather-related groundings. The airframe was pressure-washed after every third day of operations with no degradation to motor assemblies, sensor windows, or electrical connectors.

Operational Workflow for Low-Light Spraying

For teams considering similar deployments, here is the workflow McAllister Farms refined over the campaign:

T-12 hours: Download weather forecast; confirm wind speed below 10 km/h during the spray window
T-2 hours: Charge batteries; pre-mix chemical tank loads; verify RTK base station link
T-30 minutes: Arrive on-site; deploy RTK base; confirm Fix rate above 95%
T-10 minutes: Run nozzle calibration self-test; verify flow rate within ±5% of target
T-0: Launch first sortie; monitor telemetry on DJI SmartFarm controller
During operations: Swap batteries and refill tanks on a 12-minute rotation cycle
Post-flight: Download flight logs; place drift-measurement cards for lab analysis
T+48 hours: Conduct multispectral verification flight over treated blocks

Common Mistakes to Avoid

Skipping pre-flight nozzle calibration in cold temperatures. Fluid viscosity changes dramatically below 10°C. The T70P's automated calibration exists for this reason—don't override it or skip the self-test to save two minutes.

Setting spray altitude too high to "cover more ground." Every additional meter of altitude above the recommended 2–3 meters increases drift exponentially. The T70P's terrain-following radar exists to maintain tight altitude adherence. Trust it.

Ignoring RTK Fix rate degradation near metal structures. Center-pivot irrigation towers, grain bins, and power line pylons introduce multipath GPS errors. Plan flight paths with minimum 10-meter clearance from large metal objects, or accept Float-level accuracy in those zones.

Operating without a post-spray multispectral verification plan. Spraying is only half the job. Without multispectral follow-up, you have no way to identify missed strips, clogged nozzles, or sub-canopy penetration failures until yield loss makes them obvious months later.

Underestimating battery logistics. At 14.2 hectares per hour, a 480-hectare campaign requires roughly 34 hours of flight time. That translates to approximately 200+ battery swaps. Plan charging infrastructure accordingly—a portable generator and a six-bay charging hub are minimum requirements for field-scale operations.

Frequently Asked Questions

Can the Agras T70P spray effectively in complete darkness?

Yes. The T70P relies on RTK positioning, millimeter-wave radar for terrain following, and binocular vision sensors with active infrared illumination for obstacle avoidance. Visible light is not required for navigation, altitude maintenance, or spray actuation. All flight parameters—including swath width, flow rate, and ground speed—are governed by GPS-derived positioning and onboard IMU data, not optical references. The McAllister Farms campaign included 17 sorties conducted before civil twilight with no measurable performance difference compared to daylight operations.

How does spray drift performance compare between the T70P's rotary nozzles and conventional flat-fan tips?

The T70P's 16 centrifugal rotary atomization nozzles produce a more uniform droplet spectrum than hydraulic flat-fan tips, with a relative span (RS) of approximately 0.8 versus 1.2–1.5 for conventional tips. This tighter distribution means fewer very-fine droplets susceptible to drift and fewer very-coarse droplets that roll off leaf surfaces. In the Kansas case study, the 2.1% off-target drift at 20 meters reflects this advantage—comparable to ground-rig performance but without any crop trampling.

What happens if the RTK signal drops to Float during a spray run?

The T70P will continue operating in RTK Float mode with reduced positional accuracy of approximately ±0.5 meters rather than the ±2 cm centimeter precision of full Fix. The drone logs all Float-mode segments in the flight record. Best practice is to review these segments post-flight using the DJI SmartFarm platform and re-spray any affected strips if the application map shows potential gaps. During the McAllister campaign, Float episodes occurred during 1.4% of total flight time, exclusively within 15 meters of center-pivot towers, and were managed through targeted re-application passes.

Ready for your own Agras T70P? Contact our team for expert consultation.