The rapid evolution of agricultural drone technology has revolutionized crop protection strategies globally. Characterized by high operational efficiency, significant labor savings, excellent terrain adaptability, and precise chemical application, these unmanned aerial systems (UAVs) are particularly transformative. Following their successful adoption in broadacre fields, the application of agricultural drones in high-value orchard crops is gaining considerable momentum. However, this shift introduces a critical challenge: managing spray drift. Orchard spraying typically requires the agricultural drone to operate at heights of 1.5 to 2.0 meters above the canopy (often 3.5-4.0 m above ground) while employing low-volume, fine-droplet sprays. This combination creates a significantly higher risk for pesticide droplets to drift away from the target area compared to ground-based applications or field crop spraying, posing potential threats to non-target environments, bystanders, and adjacent crops. Therefore, developing robust, standardized methods to quantify and understand the drift characteristics of agricultural drone orchard sprays is not just an academic exercise but a pressing necessity for sustainable and responsible adoption of this technology.
While substantial research exists on the spray drift of agricultural drones in field crops and the simulation of their rotor downwash airflow, studies focusing on orchard applications remain in a nascent stage. Existing work often compares a single drone model with traditional ground sprayers or focuses primarily on canopy deposition and penetration. A comprehensive evaluation method that captures the full spectrum of drift—including ground sediment, near-ground, and airborne drift—for different agricultural drone platforms and nozzle types is conspicuously absent. Furthermore, conducting drift tests at the edge of real orchards is logistically challenging due to the difficulty of finding large, unobstructed areas downwind. To address these gaps, this article presents a novel, integrated testing methodology centered around an artificial orchard test bench. This method employs a combination of specialized collectors to measure canopy deposition, ground drift, near-ground drift, and, innovatively, vertical airborne drift profiles. We apply this method to evaluate and compare the spray performance and drift potential of four representative types of agricultural drones equipped with two distinct nozzle technologies.
1. Integrated Methodology for Drift Assessment
The core of this methodology is the artificial vineyard test bench, designed to simulate a typical orchard spray scenario in an open, controllable field setting. The bench was constructed using a grid of stainless steel frames (2.0 m tall, 0.5 m wide, 2.0 m long) covered with black anti-aging mesh to replicate vine foliage. Rows were spaced 2.5 meters apart. This setup allows for precise and repeatable placement of sampling apparatus without the constraints of a living orchard edge. The comprehensive sampling system deployed downwind consists of four complementary components:
- Canopy Deposition Collectors: Polyvinyl chloride (PVC) cards were clipped to the top of the artificial canopy within the swath to measure the amount and uniformity of spray directly deposited on the target.
- Ground Sediment Drift Collectors: Petri dishes were placed on level ground at distances of 3, 5, 10, 15, and 20 meters downwind from the spray swath edge to collect droplets that settle out of the air.
- Near-Ground Drift Collector (Field Drift Test Bench): A specialized, horizontally-mounted bench with pneumatically-controlled sliding covers collected droplets at a height of 0.3 meters at 0.5-meter intervals up to 10 meters downwind. This provides a high-resolution profile of drift close to the ground.
- Airborne Drift Frame Collectors: A novel apparatus consisting of vertical frames (5.5 m tall) fitted with horizontal polyethylene tubes at 0.5-meter vertical intervals. Positioned 2 meters downwind, these collectors capture the vertical distribution of drifting droplets in the air, a dimension often missed in standard tests.
2. Experimental Configuration and Agricultural Drone Platforms
The experiment evaluated four commercially prevalent types of agricultural drones, each equipped with two different nozzle types. A fluorescent tracer (Pyranine) was used in the spray solution for highly sensitive quantitative analysis. Key operational parameters were kept constant: a flight speed of 2.0 m/s and a height of 3.5 m above ground (1.5 m above the artificial canopy). Environmental conditions during tests were within acceptable ranges for drift studies (crosswind: 2.2–3.6 m/s). The tested platforms and nozzles are summarized below.
| Drone Type | Configuration | Spray System Layout | Boom Width / Span (m) | Tank Volume (L) |
|---|---|---|---|---|
| Single-Rotor (Oil-powered) | Helicopter | Boom-mounted (3 nozzles) | 1.30 | 12 |
| Hexa-rotor (Electric) | 6 rotors | Boom-mounted (4 nozzles) | 1.80 | 10 |
| Octo-rotor A (Electric) | 8 rotors | Boom-mounted (6 nozzles) | 2.40 | 20 |
| Octo-rotor B (Electric) | 8 rotors | Under-rotor mounted (4 nozzles) | 1.46 (rotor span) | 10 |
| Nozzle Model | Type | Volume Median Diameter (VMD, μm) | Portion < 75 μm (%) |
|---|---|---|---|
| TR 80-0067 | Hollow Cone | 114.9 | 16.1 |
| IDK 120-015 | Air Induction Flat Fan | 312.6 | 1.8 |
3. Data Analysis and Drift Indices
Deposition and drift rates (\( \beta_{dep} \)) were calculated from the recovered fluorescent tracer. Several indices were used to analyze the data comprehensively:
Deposition/Drift Rate & Uniformity:
$$ \beta_{dep} = \frac{(\rho_{smpl} – \rho_{blk}) \cdot F_{cal} \cdot V_{dil}}{\rho_{spray} \cdot A_{col}} $$
where \( \rho \) represents fluorescence values, \( F_{cal} \) is a calibration factor, \( V_{dil} \) is eluent volume, and \( A_{col} \) is collector area. The Coefficient of Variation (CV) was used to assess the uniformity of deposition on the canopy.
$$ CV = \frac{S}{\bar{X}} \times 100\% $$
where \( S \) is the standard deviation and \( \bar{X} \) is the mean deposition rate.
Ground Drift Indices: The Average Median Drift Rate (AMDR) was calculated from ground petri dishes to minimize outlier effects.
$$ AMDR = \frac{1}{n} \sum_{i=1}^{n} M_{\beta_{dep\%}, i} $$
where \( M_{\beta_{dep\%}, i} \) is the median drift rate at downwind distance \( i \). The downwind drift decay was modeled using an exponential function:
$$ \beta_{dep\%}(x) = a \cdot e^{b \cdot x} $$
where \( x \) is downwind distance and \( a, b \) are regression constants. The distance at which 90% of the total measurable drift has been deposited (\( x_{90\%} \)) was determined as a key buffer zone indicator.
Airborne Drift Index (ADX): A novel index was introduced to quantify the spatial drift profile captured by the vertical frames. It combines the relative drift mass and its center of mass in the vertical plane.
$$ ADX = V_r \cdot h_r $$
$$ V_r = s \cdot \sum_{i=1}^{n_p} \beta_{dep\%, i} $$
$$ h_r = \frac{ \sum_{i=1}^{n_p} (s \cdot \beta_{dep\%, i} \cdot h_i) }{h_{max}} $$
Here, \( V_r \) is the relative drift volume per unit width, \( h_r \) is the relative characteristic height (center of mass), \( s \) is the vertical spacing between collectors (0.5 m), \( h_i \) is the height of collector \( i \), and \( h_{max} \) is the maximum collection height (5.0 m). A higher ADX indicates greater and/or higher-elevation airborne drift.
4. Results and Comparative Analysis
The integrated methodology provided a multi-faceted comparison of the eight test configurations (4 drones × 2 nozzles).
Canopy Deposition: The air induction (IDK) nozzle consistently produced significantly higher and more uniform deposition on the artificial canopy compared to the hollow cone (TR) nozzle for all agricultural drone types. The Octo-rotor B model, with its under-rotor nozzle placement, showed the lowest deposition and highest variability, suggesting its optimal operational height may need adjustment.
| Drone Type | Nozzle | Avg. Deposition Rate (%) | Deposition CV (%) |
|---|---|---|---|
| Single-Rotor | IDK (Air Induction) | 72.0 | 31.2 |
| Single-Rotor | TR (Hollow Cone) | 44.5 | 87.4 |
| Hexa-rotor | IDK | 66.6 | 39.9 |
| Hexa-rotor | TR | 49.6 | 66.7 |
| Octo-rotor A | IDK | 59.7 | 42.8 |
| Octo-rotor A | TR | 52.8 | 54.7 |
| Octo-rotor B | IDK | 39.8 | 49.7 |
| Octo-rotor B | TR | 29.5 | 62.1 |
Ground and Near-Ground Drift: The exponential decay model fit the ground and near-ground drift data exceptionally well (R² > 0.92). The air induction nozzle drastically reduced downwind drift levels across all distances, despite being operated at higher application volumes. The calculated \( x_{90\%} \) values for ground drift ranged from 4.9 to 11.4 meters, suggesting a minimum buffer zone of 15 meters for vineyard spraying under the tested conditions. No statistically significant difference in ground drift potential (AMDR) was found between different agricultural drone types when using the same nozzle.
| Drone Type | Nozzle | AMDR (%) | 90% Cumulative Drift Distance, x90% (m) |
|---|---|---|---|
| Single-Rotor | IDK | 0.99 | 5.7 |
| Single-Rotor | TR | 2.26 | 10.7 |
| Hexa-rotor | IDK | 1.06 | 4.9 |
| Hexa-rotor | TR | 2.12 | 8.6 |
| Octo-rotor A | IDK | 0.85 | 11.4 |
| Octo-rotor A | TR | 2.01 | 9.4 |
| Octo-rotor B | IDK | 1.74 | 6.2 |
| Octo-rotor B | TR | 3.35 | 9.8 |
Airborne Drift Profiles: The vertical frame collectors revealed critical insights. Drift rates were highest near the ground and decreased with height. Notably, for the hollow cone nozzle, a substantial portion of collected drift (up to 16.5%) was found above the flight height (3.5 m), indicating droplets caught in the complex vortex systems generated by the interaction of rotor downwash and crosswind. These droplets are prone to long-range atmospheric transport and evaporation. The Airborne Drift Index (ADX) successfully quantified these profiles, with the hollow cone nozzle producing significantly higher ADX values than the air induction nozzle for each agricultural drone.
| Drone Type | Nozzle | Airborne Drift Index (ADX) |
|---|---|---|
| Single-Rotor | IDK | 6.3 |
| Single-Rotor | TR | 21.8 |
| Hexa-rotor | IDK | 3.6 |
| Hexa-rotor | TR | 12.6 |
| Octo-rotor A | IDK | 5.4 |
| Octo-rotor A | TR | 13.9 |
| Octo-rotor B | IDK | 12.5 |
| Octo-rotor B | TR | 19.2 |
5. Correlation of Methods and Practical Implications
A key finding was the strong correlation between the results from different sampling methods. The ground drift AMDR showed significant statistical correlations with canopy deposition rate (negative), canopy deposition CV (positive), near-ground AADR (positive), and the ADX (positive). This validates that simpler or more targeted measurements (like canopy deposition uniformity or a single vertical frame) can provide reliable indicators of overall drift risk for an agricultural drone configuration, streamlining future testing protocols.
The analysis underscores several critical points for the operation and development of orchard-specialized agricultural drones:
- Nozzle Selection is Paramount: The air induction nozzle (IDK) dramatically improved canopy deposition, uniformity, and reduced all forms of drift compared to the conventional hollow cone nozzle, demonstrating that nozzle technology is a more decisive factor than the agricultural drone airframe type itself for mitigating drift in orchard settings.
- Vortex-Driven Drift: The collection of drift above flight altitude highlights the significant role of rotor-induced vortices in transporting fine droplets. Stronger downwash does not inherently guarantee less drift; it can contribute to more energetic and complex drift-prone airflow patterns.
- Buffer Zone Recommendation: Based on the \( x_{90\%} \) measurements, a downwind buffer zone of at least 15 meters should be considered for agricultural drone vineyard spraying under moderate wind conditions.
- Utility of the ADX: The Airborne Drift Index provides a single, quantitative metric to compare the spatial drift profile of different agricultural drone and nozzle combinations, useful for manufacturers and regulators.
6. Conclusion
This study presents and validates a comprehensive, field-based methodology for assessing the spray drift of agricultural drones in orchard applications. By employing an artificial orchard test bench coupled with a multi-layered sampling strategy—capturing canopy deposition, ground sediment, near-ground, and vertical airborne drift—the method provides a holistic view of spray performance and off-target loss. The application of this method to four common agricultural drone types revealed that while drone platform differences were less pronounced, the choice of nozzle technology had a profound impact. Air induction nozzles significantly outperformed conventional hollow cone nozzles in enhancing target deposition and reducing drift. The introduced Airborne Drift Index (ADX) offers a valuable tool for quantifying spatial drift. Furthermore, the strong correlations between different measurement techniques suggest pathways for simplifying future standard test protocols. This work provides essential data and methodological frameworks to support the responsible development, regulation, and operation of agricultural drones in sensitive orchard ecosystems, ensuring that the benefits of this transformative technology are realized while minimizing its environmental footprint.