Comparative analysis of operational capacity and acquisition costs of manned agricultural aircraft and spraying drones

Abstract

Aerial spraying is a fundamental strategy for maintaining operational efficiency and productivity in Brazilian agriculture. In recent years, the rapid expansion of remotely piloted aircraft (drones) has intensified the debate over their technical, operational, and economic performance compared with traditional manned agricultural aircraft. In this context, this study developed a mathematical model to evaluate and compare the operational capacity of manned agricultural aircraft (EMB-202 and EMB-203) and unmanned aircraft (AGRAS T10, T20P, T30, and T40) under three application rates: 5, 10, and 20 L ha⁻¹. Operational data were obtained through direct inquiries with 41 aircraft maintenance facilities certified by the National Civil Aviation Agency (ANAC) and 41 agricultural drone dealerships accredited by manufacturers across Brazil's major agricultural regions. The results indicated that five AGRAS T40 drones operating simultaneously can achieve a daily operational capacity comparable to that of the EMB-203 at an application rate of 5 L ha⁻¹, reaching approximately 900 ha day⁻¹ while requiring an acquisition cost approximately 3.6 times lower than that of a new manned aircraft. Kernel density estimation revealed a greater concentration of manned aircraft in the state of Mato Grosso, whereas unmanned aircraft were more concentrated in the state of São Paulo. These findings indicate that agricultural drones are a competitive alternative to manned aircraft in terms of operational capacity and acquisition costs for medium-sized farms, with strong potential for technological and operational expansion into other agricultural regions of Brazil, particularly the Northeast.

Keywords:
remotely piloted aircraft; agricultural aviation; application technology; operational capacity; precision agriculture; spatial analysis

Introduction

Aerial application plays a strategic role in Brazilian agriculture, particularly given the vast land area dedicated to crop production. This technology enables efficient, large-scale control of pests and diseases, contributing to the maintenance of crop productivity (Baio et al., 2019).

The use of pesticides aims to manage weeds, pathogens, and pests while ensuring agronomic efficiency without compromising the safety of operators, consumers, or the environment (Carvalho & Oliveira, 2019). However, problems such as inadequate deposition, spray drift, and non-uniform spray distribution still occur when technical planning is insufficient or application technology is improperly implemented (Antuniassi & Boller, 2019). Application efficiency is a key determinant of the success of integrated pest management (Oliveira, 2021) and of improving operational quality in modern agriculture (Souza, 2020).

In this context, aerial application has become a vital component of Brazilian agriculture (De Miranda, 2018), alongside the expansion of agricultural drones (Wang & Zhang, 2021; Massey et al., 2018). Properly defining spray swath width and ensuring application uniformity are critical to operational efficiency (Carvalho & Cunha, 2019). Mathematical modeling therefore serves as an essential tool for evaluating the operational performance and economic feasibility of these technologies (Luenberger, 2016; Carvalho et al., 2021).

The aerial application sector continues to evolve through advances in operational safety (Mwitta & Rains, 2024), energy efficiency (Brelje & Martins, 2019), and technological innovation (Silva & Henkes, 2025). In addition to technical aspects, regulatory (ANAC, 2023), environmental, and market factors directly influence the adoption of these technologies (Lan et al., 2017). Economic analysis also plays a key role in optimizing operational performance and supporting strategic decision-making (Melo, 2020; Pacheco, 2000; Schimmelpfennig, 2016).

Despite the concurrent expansion of manned agricultural aircraft and spraying drones, comprehensive studies comparing their operational capacity and acquisition costs under different application requirements remain limited. This study considered three operational scenarios 5, 10, and 20 L ha⁻¹ representing distinct application volumes that directly affect operational capacity (ha h⁻¹), the number of units required, and acquisition costs. Accordingly, this study developed a mathematical model to compare the operational capacity and application costs of manned agricultural aircraft and spraying drones across these scenarios, providing a technical basis for planning and managing aerial applications at different production scales. As part of the results, graphical visualizations were generated to provide a clear and informative understanding of the operational capacity of the different aircraft models.

Material and Methods

Selection of Aircraft and Data Collection

For this study, the EMB-202 and EMB-203 agricultural aircraft models were selected because they account for 21% of the registered agricultural aviation fleet in Brazil (ANAC, 2024). Although production of the EMB-202 has ceased, 391 units remain in operation. The EMB-203, which is currently in production, features an ethanol-powered engine, compatibility with more than two factory-installed DGPS (Differential Global Positioning System) models, a laser altimeter, and a VHF (Very High Frequency) radio. Its active fleet comprises 181 aircraft. Supplementary information on the market and equipment imports was obtained from official foreign trade databases (SISCOMEX, 2024).

Operational data were collected through direct inquiries with 41 aircraft maintenance facilities certified by the National Civil Aviation Agency (ANAC) and 41 manufacturer-accredited agricultural drone dealerships distributed across Brazil's major agricultural production regions. Data collection followed a structured protocol using standardized forms administered between March and September 2023. The collected data were validated by cross-referencing them with available technical records and the technical specifications provided in the manufacturers' operational manuals. Complete technical specifications for the unmanned aircraft models are presented in Tables 5 and 6, including tank capacity, rotor span, operating speed, operating flight altitude, and atomization system type.

Table 1 presents the geographic distribution of EMB-202 and EMB-203 agricultural aircraft by Brazilian state, together with the corresponding certified maintenance facilities. This information provides a technical overview of the distribution of both the aircraft fleet and the associated certified maintenance infrastructure. Consequently, it supports a comprehensive assessment of the infrastructure available for these agricultural aircraft throughout Brazil.

Table 1
Distribution of the 391 EMB-202 Aircraft, 181 EMB-203 Aircraft, and 41 Certified Maintenance Facilities across Brazilian States.

State	EMB-202	EMB-203	Maintenance Facility
AL	2	3	1
AM	1
BA	10	13	2
GO	27	16	4
MA	2	10	2
MG	6	14
MS	26	9	2
MT	126	64	9
PA	19	7	1
PE	2		1
PI	2	3	1
PR	19	8	2
RJ			1
RO	2	6
RR	1	2
RS	87	6	5
SE		1
SP	33	14	8
TO	26	5	2
Total	391	181	41

Table 2 presents key information on the market and technical specifications of the EMB-202 and EMB-203 models. The table includes variables such as the distribution of service providers, average aircraft acquisition cost, engine specifications, fuel costs, wingspan, and payload capacity.

Table 2
Market analysis and technical specifications of the EMB-202 and EMB-203 models for decision-making under different operational scenarios.

Model	Corporate Ownership	Private Ownership	Aircraft Acquisition Cost	Engine	Fuel Cost	Wingspan	Payload Capacity
	(%)	(%)	(R$)	(um)	(R$)	(m)	(L)
EMB-202	70	30	987 mil	Avgas	8.62	11.0	950
EMB-203	37	63	3.8 M	Ethanol	3.56	13.3	1050

• * Published data from ANAC (2024) and ANAC-certified maintenance facilities.

For a detailed analysis of operational performance, Table 3 presents application-specific data, including average spray swath width, operating speed, and operational capacity (ha h⁻¹) at three application rates (5, 10, and 20 L ha⁻¹). The average operational capacity values were compiled from data provided by authorized service centers.

Table 3
Technical specifications of the EMB-202 and EMB-203 models.

Model	Spray Swath Width	Operating Speed	Operational Capacity 5 L ha-1	Operational Capacity 10 L ha-1	Operational Capacity 20 L ha-1
	(m)	(Km h-1)	(ha h-1)	(ha h-1)	(ha h-1)
EMB-202	18.6	176	140	100	60
EMB-203	21.6	185	180	130	80

• * Spray swath width values were obtained from Althman (2021), whereas operational capacity data were collected from ANAC-certified maintenance facilities.

Regarding spraying drones, multirotor aircraft predominate, accounting for 77% of the total fleet. This market segment comprises 582 DJI AGRAS T10 units, 478 DJI AGRAS T20P units, 362 DJI AGRAS T30 units, and 1,343 DJI AGRAS T40 units. In addition, 41 authorized dealerships provided the operational data used in this study (Kern, 2023). These figures illustrate the widespread adoption of multirotor spraying drones in Brazilian agriculture, consistent with the findings of Silva & Henkes (2025).

Table 4 presents the geographic distribution of spraying drones based on data compiled from the ANAC database (2024). The records show the distribution of these models across Brazilian states, focusing exclusively on their geographic distribution. Operational data were obtained separately from authorized dealerships.

Table 4
State-level distribution of AGRAS T10, AGRAS T20P, AGRAS T30, and AGRAS T40 spraying drones and authorized dealerships.

State	AGRAS T10	AGRAS T20P	AGRAS T30	AGRAS T40	Dealership
AC	1	2		5	1
AL	5	4	3	1
AM		2	2	5
AP				1
BA	14	8	15	31	3
CE	2	1		15
DF	3	7	5	5	1
ES	59	27	13	34
GO	30	30	21	85	2
MA	5	2	1	22
MG	83	76	48	167	1
MS	10	6	6	70
MT	19	15	12	111	2
PA	22	10	23	49	1
PB	2	2	2	4
PE	22	7	5	5
PI	2		1	2
PR	54	55	52	166	4
RJ	8	7	1	4
RN			1	3
RO	12	13	10	46
RR	1			3
RS	59	39	38	124	20
SC	46	39	22	117	2
SE	1	7	5	4	1
SP	114	109	74	251	2
TO	8	10	2	13	1
Total	582	478	362	1343	41

• *Published data from ANAC (2024) and information provided by authorized dealerships (Kern, 2023).

Equipment Technical Specifications

Table 5 presents market information and technical specifications for the evaluated models. The table includes the distribution of service providers, average acquisition cost, and the costs of the battery, charger, and generator for each model and its respective payload capacity.

Table 5
Market information and technical specifications of AGRAS T10, AGRAS T20P, AGRAS T30, and AGRAS T40 spraying drones, including market share among service providers, average acquisition cost, battery cost, charger and generator cost, and payload capacity.

Model	Corporate Ownership	Private Ownership	Drone Acquisition Cost	Battery Cost	Charger + Generator	Payload Capacity
	(%)	(%)	(R$)	(R$)	(R$)	(L)
AGRAS T10	50	50	50 mil	10 mil	20 mil	10
AGRAS T20P	37	63	85 mil	12 mil	23 mil	20
AGRAS T30	52	48	100 mil	17 mil	25 mil	30
AGRAS T40	32	68	115 mil	20 mil	36 mil	40

• * Information collected from the 41 authorized dealerships (Kern, 2023).

For a comprehensive analysis of operational performance, Table 6 presents application-specific data for the evaluated models, including average spray swath width, operating speed, and operational capacity (ha h⁻¹) at three application rates. Spray swath width was determined assuming a flight altitude of 3 m for the EMB-202 and EMB-203 aircraft, according to the Embraer Operations Manual (2020), and 2–3 m for the drones, based on the DJI Technical Specifications (2024). The reported values represent the effective spray swath width, assuming an average 20% overlap between adjacent passes to ensure uniform spray deposition.

Table 6
Market data and technical specifications of AGRAS T10, AGRAS T20P, AGRAS T30, and AGRAS T40 spraying drones.

Model	Spray Swath Width	Operating Altitude	Operating Speed	Operational Capacity 5 L ha-1	Operational Capacity 10 L ha-1	Operational Capacity 20 L ha-1
	(m)	(m)	(Km.h-1)	(ha h-1)	(ha h-1)	(ha h-1)
AGRAS T10	5.5	3	18	7	5.0	3.5
AGRAS T20P	6.5	3	20	12	7.5	5.5
AGRAS T30	9.0	3	22	16	11.0	7.0
AGRAS T40	10.0	4	25	22	16.0	10.0

• * Information collected from the 41 authorized dealerships (Kern, 2023).

In summary, the information presented in Tables 2, 3, 5, and 6 provides a comprehensive overview of the technical specifications and operational performance of the EMB-202 and EMB-203 aircraft, as well as the evaluated spraying drone models. Together, these data characterize the technical, operational, and economic parameters for selecting the most appropriate strategy for different agricultural scenarios.

Operational Parameters

The optimization model comparing manned agricultural aircraft and spraying drones was based on three operational parameters: application rate, operational capacity (ha h⁻¹), and daily operating time. The application rate represents the volume of spray solution applied per unit area, operational capacity represents the area covered per hour, and daily operating time defines the available operating period for each application platform.

Application rates of 5, 10, and 20 L ha⁻¹ were selected because they represent the range commonly recommended for pesticide applications in Brazil. The 5 L ha⁻¹ rate corresponds to ultra-low-volume aerial applications (Antuniassi & Boller, 2019); 10 L ha⁻¹ represents an intermediate application rate selected to represent agricultural spraying operations (Nascimento et al., 2018); and 20 L ha⁻¹ corresponds to applications requiring higher spray volumes according to crop, product, or phenological stage requirements (Cunha et al., 2017). Evaluating these three application rates enabled comparison of the operational performance of aircraft and drones under different spraying scenarios.

Operational capacity (ha h⁻¹) was determined for each aircraft and drone model based on operational parameters obtained through collaboration with 41 aircraft maintenance facilities and 41 authorized drone dealerships, representing a wide range of operating conditions across Brazil. The dataset included spray swath width, operating speed, payload capacity, operational capacity (ha h⁻¹), and average daily operating time. These parameters enable comparisons of the operational performance of the evaluated aircraft in pesticide applications and support operational planning and decision-making in precision agriculture (Wang & Zhang, 2021).

Daily operating time, a critical element for the planning and execution of agricultural activities, was established as a fixed parameter in the model. Based on typical operating conditions, values of 480 min day⁻¹ were adopted for spraying drones (including night operations) and 300 min day⁻¹ for manned aircraft. This difference reflects the operational limitations of manned aircraft, which generally perform spraying operations only during daylight hours because of safety and visibility requirements. Spraying drones, however, can also operate at night, extending the available daily operating period. Consequently, nighttime operation increases the time available for spraying, providing greater operational flexibility and improving the use of favorable weather conditions. These assumptions provide a realistic representation of the operational differences between manned aircraft and spraying drones.

Data Processing and Analysis

Data analysis was conducted using the Python programming language, version 3.10.12 (Van Rossum & Drake, 2009). The Pandas v2.0.3 (McKinney, 2010) and NumPy v1.24.3 (Harris et al., 2020) libraries were used for data manipulation and analysis, while Matplotlib v3.7.1 (Hunter, 2007) and Seaborn v0.12.2 (Waskom, 2021) were employed for data visualization. These tools were used to develop the mathematical model and generate the visualizations for the comparative analysis of agricultural aircraft and spraying drones.

The data preparation phase plays a crucial role in constructing the mathematical model by ensuring data integrity and quality, as highlighted by Luenberger (2016). The collected data were initially entered into a Google Spreadsheet and subsequently exported as a .csv file to facilitate data manipulation. The Pandas library within the Python environment was used to import and manipulate the data, ensuring an efficient workflow while preserving data integrity.

During this phase, a thorough review of the datasets was conducted to identify and address inconsistencies, missing values, and potential outliers. The following issues were identified: (i) three duplicate aircraft records in the ANAC database, which were removed; (ii) twelve pieces of equipment with incomplete technical specifications, for which the missing information was verified and supplemented through consultation with the manufacturers; (iii) discrepancies in eight records related to spray swath width, which were corrected based on official technical manuals; and (iv) five outliers associated with operational time, which were analyzed individually and retained after confirmation with the operating companies.

Following this stage, normalization and standardization techniques were applied to ensure scale uniformity and facilitate model convergence. Preprocessing strategies were employed to address data heterogeneity, including variable standardization, handling of missing values, and outlier detection. Analyses were conducted using the Pandas, NumPy, and Scikit-learn libraries within the Python environment. Descriptive statistical analyses, including the mean, standard deviation, and quartiles, were performed to characterize the dataset. In addition, data distribution was visualized using Kernel Density Estimation (KDE) with a Gaussian kernel and bandwidth estimated according to Scott's rule (Scott, 1992). Spatial analysis of the geographic distribution of the facilities was performed using interpolation and spatial smoothing techniques based on KDE, enabling the identification of concentration patterns and regional distributions.

Mathematical Modeling

To estimate the daily operational capacity of the aircraft and drones evaluated in this study, an equation relating effective operating time to average operational capacity was used. This relationship allows the calculation of the total area that each model can cover per day under average operating conditions. The equation used to estimate daily operational capacity is expressed as:

$\text{OpCap}{}_{\text{daily}}=\text{Effective time}\text{×}\text{Average hectares}$ (1)

Where:

OpCap_daily: daily operational capacity;

Effective Time: effective daily operating time of the aircraft (in minutes);

Average hectares: average area covered (in hectares per hour).

The effective operating time considered was 300 min for manned aircraft and 480 min for spraying drones. These values were derived from an analysis of 156 flight plans provided by 23 aerial spraying service companies between March and September 2023. The data were processed using spreadsheets to determine average operating times and subsequently validated against operational information provided by the National Civil Aviation Agency (ANAC).

The adopted model is comparative and simplified, providing a common basis for evaluating the different application platforms. Time associated with refueling, supply logistics, internal transit, battery replacement, and routine maintenance was incorporated indirectly into the average operational capacity values obtained from the participating companies rather than modeled individually.

This calculation provides a quantitative estimate of the operational performance of each evaluated model. By combining effective operating time with average operational capacity, the model estimates daily application capacity, an important indicator for operational planning and large-scale pesticide application.

To facilitate comparisons among application rates, the dataset was divided into three groups corresponding to application rates of 5, 10, and 20 L ha⁻¹. This segmentation enabled the evaluation of model performance under different application scenarios and the identification of operational differences associated with each application rate.

To evaluate operational capacity under different production scenarios, a simulation was performed for application areas ranging from 0 to 1,000 ha, with increments of 2 ha. This range was selected because it encompasses most Brazilian farms that use aerial application technologies, allowing the evaluation of different operational scales and the identification of transition points at which the use of spraying drones or manned agricultural aircraft becomes operationally advantageous. This approach enabled a detailed comparison of model performance under different operating conditions.

The simulation estimated the number of days, operating hours, and aircraft or drones required for each scenario using the following equations:

${\text{Days}}_{\text{required}}=\frac{\text{Area}}{{\text{opCap}}_{\text{daily}}}$ (2)

${\text{Hours}}_{{\text{required}}^{-}}={\text{Days}}_{\text{required}}\times \text{Effective Time}$ (3)

${\text{Aircraft}}_{\text{required}}=\frac{{\text{Hours}}_{\text{required}}}{\text{Effective Time}}$ (4)

Where:

Days_daily: number of days required to treat the specified area;

Hours_required: total operating hours required to treat the specified area;

Aircraft_required: number of aircraft or drones required to complete the application within the projected time frame.

This simulation provides a comprehensive assessment of operational viability under different production scales and supports planning and decision-making for large-scale agricultural pesticide applications across diverse farming areas.

Spatial and Statistical Analysis

Spatial analysis was conducted using the geographic records of agricultural aircraft and spraying drones through Kernel Density Estimation (KDE). Location data were obtained from official records of the National Civil Aviation Agency (ANAC) and organized to generate thematic maps. Cartographic visualization and spatial interpolation were performed using QGIS software, enabling the identification of concentration patterns and the regional distribution of the equipment. Kernel Density Estimation was used as a spatial smoothing technique to identify areas with the highest concentration of the analyzed technologies. In addition, a comparative analysis was conducted between equipment distribution and rural property size using data from the Land Management System (SIGEF).

This analysis allowed the assessment of the relationship between the adoption of spraying technologies and the land tenure profile of the analyzed regions. Descriptive statistical analyses were performed using the Pandas and NumPy libraries within the Python environment, whereas spatial representation and density analyses were conducted in a GIS environment. Together, these complementary approaches supported the operational and economic comparison between manned agricultural aircraft and spraying drones.

Results and Discussion

The visualizations reveal clear differences in operational capacity between manned agricultural aircraft and spraying drones. In general, manned aircraft exhibit greater daily operational capacity when evaluated individually because of their higher operating speeds and wider spray swath widths. However, when the number of units required to treat a given area per day is considered, multiple drones can achieve comparable performance, highlighting the operational flexibility of this technology.

A detailed analysis of the results for daily spraying capacity at 5 L ha⁻¹ an application rate associated with a wider spray swath indicates the potential for greater operational capacity and lower operating costs (Carvalho et al., 2021). This application rate is particularly relevant in situations where application precision is critical because of the risk of spray drift.

Effective spray swath width and flight-line spacing are distinct operational parameters. The effective spray swath width represents the actual width over which adequate product deposition on the target is achieved, whereas flight-line spacing refers to the operational distance between successive passes of the aircraft or drone. In practice, this spacing can be adjusted to include a safety overlap between adjacent swaths, ensuring uniform coverage (Wang et al., 2019).

Weather conditions, including wind speed, temperature, and relative humidity, directly influence spray drift and, consequently, the effective spray swath width. Under favorable application conditions (wind speed below 5 km h⁻¹, temperature between 20 and 25 °C, and relative humidity between 60% and 80%), the effective spray swath width generally approaches the theoretical working width. However, under less favorable environmental conditions, flight-line spacing should be reduced by approximately 15% to 25% to maintain application uniformity (Carvalho et al., 2021). Figure 1 illustrates the operational capacity of the evaluated models studied.

Figure 1
Daily Application Capacity (ha day⁻1) of Manned Agricultural Aircraft (EMB-202 and EMB-203) and Spraying Drones (AGRAS T10, AGRAS T20P, AGRAS T30, and AGRAS T40) at an Application Rate of 5 L ha⁻1. X-Axis: Application Area (ha); Y-Axis: Number of Aircraft or Drones Required to Complete the Application Within One Working Day.

Five AGRAS T40 drones, seven AGRAS T30 drones, or ten AGRAS T20P drones are required to match the EMB-203 operational capacity of 900 ha day⁻¹. Similarly, four AGRAS T40 drones, six AGRAS T30 drones, or eight AGRAS T20P drones are required to match the EMB-202 operational capacity of 700 ha day⁻¹. These differences are directly related to the operational characteristics of each platform. Manned agricultural aircraft operate at higher speeds and have wider spray swath widths, resulting in greater daily operational capacity per unit. In contrast, spraying drones have lower payload capacities and narrower spray swath widths, requiring multiple units to treat equivalent areas. However, the ability to operate several drones simultaneously may provide an operational advantage in smaller fields or areas with greater spatial fragmentation. The corresponding acquisition costs are presented in Table 7.

Table 7
Comparison of Acquisition Costs Between EMB-203 and EMB-202 Aircraft and AGRAS Spraying Drones at an Application Rate of 5 L ha⁻1.

Model	Units	Aircraft Acquisition Cost	Model	Units	Aircraft Acquisition Cost
	(um)	(R$)		(um)	(R$)
EMB-203	1	3,800,000.00	EMB-202	1	987,000.00
AGRAS T40	5	1,055,000.00	AGRAS T40	4	844,000.00
AGRAS T30	7	1,237,500.00	AGRAS T30	6	1,056,000.00
AGRAS T20P	10	1,440,000.00	AGRAS T20P	8	1,152,000.00
AGRAS T10	16	1,632,000.00	AGRAS T10	13	1,326,000.00

• * Survey data based on a standard multirotor system (one drone, three batteries, one charger, and one generator).

The application rate of 10 L ha⁻¹ is commonly adopted for both aircraft and is widely used for the precise and effective application of agricultural inputs. Accordingly, a detailed analysis of the results presented in Figure 2 is warranted.

Figure 2
Daily application capacity (ha day⁻1) of manned agricultural aircraft (EMB-202 and EMB-203) and spraying drones (AGRAS T10, AGRAS T20P, AGRAS T30, and AGRAS T40) at an application rate of 10 L ha⁻1. X-axis: application area (ha); Y-axis: number of aircraft or drones required to complete the application within one working day.

Five AGRAS T40 drones, eight AGRAS T30 drones, or eleven AGRAS T20P drones are required to match the EMB-203 operational capacity of 650 ha day⁻¹. Similarly, four AGRAS T40 drones, six AGRAS T30 drones, or nine AGRAS T20P drones are required to match the EMB-202 operational capacity of 500 ha day⁻¹. The corresponding acquisition costs are presented in Table 8.

Table 8
Acquisition Cost Comparison Between EMB-202 and EMB-203 Aircraft and AGRAS Spraying Drones at an Application Rate of 10 L ha⁻1.

Model	Units	Aircraft Acquisition Cost	Model	Units	Aircraft Acquisition Cost
	(um)	(R$)		(um)	(R$)
EMB-203	1	3,800,000.00	EMB-202	1	987,000.00
AGRAS T40	5	1,055,000.00	AGRAS T40	4	844,000.00
AGRAS T30	8	1,408,000.00	AGRAS T30	6	1,056,000.00
AGRAS T20P	11	1,584,000.00	AGRAS T20P	9	1,296,000.00
AGRAS T10	17	1,734,000.00	AGRAS T10	13	1,326,000.00

• * Survey data based on a standard multirotor system (one drone, three batteries, one charger, and one generator).

No significant variation in drone acquisition costs was seen compared with the application scenario using a rate of 5 L ha⁻¹. This behavior is associated with the reduction in the operational capacity of manned aircraft as the application rate increases, partially reducing the operational advantage of these platforms. Consequently, at intermediate application rates, such as 10 L ha⁻¹, the choice between manned aircraft and spraying drones tends to depend not only on operational capacity but also on logistical factors, property size, and the availability of supporting infrastructure. Only the AGRAS T40 maintained the same acquisition cost, while the acquisition costs of the AGRAS T30, AGRAS T20P, and AGRAS T10 once again exceeded those of the used manned aircraft. The following analysis focuses on a daily application rate of 20 L ha⁻¹, which stands for a more demanding operating scenario for both application platforms.

Figure 3 illustrates these variations, enabling a more in-depth analysis of model performance under high application-rate conditions compared with the other application rates. Examining these results makes it possible to identify the models that provide the most consistent and efficient application performance at an application rate of 20 L ha⁻¹.

Figure 3
Daily application capacity (ha day⁻1) of manned agricultural aircraft (EMB-202 and EMB-203) and spraying drones (AGRAS T10, AGRAS T20P, AGRAS T30, and AGRAS T40) at an application rate of 20 L ha⁻1. X-axis: application area (ha); Y-axis: number of aircraft or drones required to complete the application within one working day.

As shown in Figure 3, five AGRAS T40 drones, seven AGRAS T30 drones, or eleven AGRAS T20P drones are required to match the EMB-203 operational capacity of 400 ha day⁻¹. Similarly, four AGRAS T40 drones, six AGRAS T30 drones, or seven AGRAS T20P drones are required to match the EMB-202 operational capacity of 300 ha day⁻¹. The corresponding acquisition costs are presented in Table 9.

Table 9
Acquisition Cost Comparison Between EMB-202 and EMB-203 Aircraft and AGRAS Spraying Drones at an Application Rate of 20 L ha⁻1.

Model	Units	Aircraft Acquisition Cost	Model	Units	Aircraft Acquisition Cost
	(um)	(R$)		(um)	(R$)
EMB-203	1	3,800,000.00	EMB-202	1	987,000.00
AGRAS T40	5	1,055,000.00	AGRAS T40	4	844,000.00
AGRAS T30	7	1,232,000.00	AGRAS T30	6	1,056,000.00
AGRAS T20P	9	1,296,000.00	AGRAS T20P	7	1,008,000.00
AGRAS T10	15	1,530,000.00	AGRAS T10	11	1,122,000.00

• * Survey data based on a standard multirotor system (one drone, three batteries, one charger, and one generator).

At the 20 L ha⁻¹ application rate, the observed behavior is similar to that at the other application rates, with relatively slight differences in the acquisition costs required to achieve equivalent operational capacities. However, higher application volumes tend to favor platforms with greater payload capacity, such as manned agricultural aircraft, because the increased spray volume per hectare reduces operational capacity, particularly for drones with smaller tanks. Currently, 70% of EMB-202 aircraft are operated by specialized aerial spraying companies, whereas 63% of EMB-203 aircraft are operated by producers on their own farms, as shown in Table 2.

Based on these findings and the data presented in Table 1, derived from ANAC (2024) records, a Kernel Density Estimation (KDE) map was generated. This spatial interpolation and smoothing technique (QGIS, 2023) provides a detailed view of the geographic distribution of these aircraft, as illustrated in Figure 4.

Figure 4
Spatial Distribution of EMB-202 and EMB-203 Aircraft Based on Kernel Density Estimation (KDE) Across Brazilian Mesoregions.

*Kernel Density Estimation (KDE) was used to visualize the geographic concentration of the evaluated aircraft. The color gradient represents density intensity (red = highest concentration; blue = lowest concentration). This method enables the identification of regions with the highest concentration of the analyzed aircraft. Operational range indicates the average operating radius from the identified base locations.

The highest concentrations of agricultural aircraft were observed in the Brazilian mesoregions of Northern Mato Grosso, Southeastern Mato Grosso, Southwestern Rio Grande do Sul, Southern Goiás, and Northeastern Mato Grosso. These regions are major production frontiers for soybeans, rice, and cotton (Antuniassi & Boller, 2019).

Spraying drones, whose adoption has increased rapidly in Brazilian agriculture, exhibit distinct operational profiles. Earlier models released (AGRAS T10 and AGRAS T30) are predominantly operated by service providers, whereas the more recent models (AGRAS T40 and AGRAS T20P) are managed by individual operators, accounting for 68% and 63% of the respective fleets. This distribution reflects the growing integration of spraying drones into precision agriculture and their adaptation to different operational demands. The resulting spatial distribution is presented in Figure 5.

Figure 5
Spatial Distribution of Spraying Drones Based on Kernel Density Estimation (KDE) Across Brazilian Mesoregions.

*Kernel Density Estimation (KDE) was used to visualize the geographic concentration of the evaluated equipment. The color gradient indicates density intensity (red = highest concentration; blue = lowest concentration). This method enables the identification of regions with the highest equipment concentration. Operational range indicates the average operating radius from the identified base locations.

High concentrations were observed in the Brazilian mesoregions of Ribeirão Preto, Triângulo Mineiro, Noroeste Rio-Grandense, Oeste Paranaense, and Centro Goiano. This spatial distribution is directly associated with the structural characteristics of Brazilian agriculture. Regions dominated by large agricultural holdings and extensive monoculture areas, such as Mato Grosso and Goiás, tend to favor the use of manned agricultural aircraft because of their high operational capacity. Conversely, regions characterized by smaller farms and greater crop diversity exhibit higher adoption rates of spraying drones, as these platforms offer greater operational flexibility and require less airport infrastructure. The Northeast region also presents considerable development potential because of the size of its rural properties and the diversity of crops grown there.

The operational profiles presented in Figures 4 and 5 highlight the importance of operational capacity in aerial spraying, as discussed by Pacheco (2000). This parameter, which reflects both application time and the area covered, directly influences operational performance and contributes to the optimization of time and resources. In addition, several geographic factors influence the selection of spraying technology, including crop type, topography, climate, and farm size.

Beyond farm size, crop composition also influences the adoption of spraying technologies. In the state of São Paulo, crops such as sugarcane and citrus predominate and are typically cultivated under more heterogeneous topographical conditions and in more fragmented production areas, which can favor the use of spraying drones. In contrast, the state of Mato Grosso is dominated by extensive soybean, maize, and cotton production distributed across large, contiguous areas, a scenario that favors the high operational capacity of manned agricultural aircraft. Figure 6, based on SIGEF (2024) records and data from the municipalities where the evaluated models operate, compares the performance of drone and aircraft models considering the predominance of drones in São Paulo and manned aircraft in Mato Grosso.

Figure 6
Comparison of the Operational Performance of Spraying Drones in São Paulo and Manned Agricultural Aircraft in Mato Grosso Across Rural Property Sizes Based on SIGEF (2024) Records.

The results indicate that the operational range of the three drone models evaluated in the state of São Paulo is generally limited to rural properties of up to approximately 550 ha, a value comparable to the average operational range of manned agricultural aircraft in the state of Mato Grosso. Notably, the analyzed properties exceed two fiscal modules, a threshold that requires 20% of the land to be maintained as permanent preservation areas. Consequently, the safety buffer required for spraying drones is approximately 12 times smaller than that required for manned aircraft, with important implications for operational planning and field management. Understanding these operational constraints is essential when evaluating the daily application capacity of different aircraft and drone models, as also reported by Carvalho et al. (2021) in their analyses of application rates.

The mathematical model developed in this study enabled a comparative assessment of daily application capacity at three application rates (5, 10, and 20 L ha⁻¹), highlighting significant differences among the evaluated technologies. Although manned agricultural aircraft exhibited greater operational capacity on an individual basis, spraying drones achieved comparable performance when operated in multiple-unit configurations. Therefore, these technologies should not be viewed as competitors but rather as complementary application platforms that can be selected according to property size, crop characteristics, and operational requirements.

The model also revealed significant differences in operational efficiency among the evaluated platforms. At an application rate of 5 L ha⁻¹, the AGRAS T40 required fewer units to match the operational capacity of the EMB-202 and EMB-203, demonstrating the operational feasibility of multi-drone applications. As shown in Tables 3 and 6, the AGRAS T40 provides an effective spray swath width approximately 46% that of the agricultural aircraft. According to ANAC (2024), Brazil currently has approximately 3.5 agricultural drones in operation for every manned agricultural aircraft. This rapid expansion, combined with advances in precision agriculture, is reshaping aerial spraying operations. The integration of real-time monitoring systems further improves application precision by enabling detailed assessments of field variability, thereby optimizing resource use and operational efficiency.

The AGRAS T40 maintained constant acquisition costs across the three application rates evaluated when compared with the EMB-203 and EMB-202 aircraft, with acquisition costs 3.6 and 1.2 times lower, respectively. Furthermore, the comparison between aircraft and drone acquisition costs revealed a favorable relationship between operational capacity and equipment cost, particularly at lower application rates.

This performance can be partially attributed to recent advances in the spraying systems used in agricultural drones. Chen et al. (2022) reported that centrifugal atomizers used on drones produce droplets with a Volume Median Diameter (VMD) ranging from 150 to 250 μm, whereas conventional hydraulic nozzles used on agricultural aircraft generate droplets with a VMD between 200 and 400 μm. This characteristic provides greater control over the droplet spectrum and allows better adaptation to different operating conditions. Although smaller droplets may be more susceptible to drift under certain conditions, drone operations are typically conducted at lower flight altitudes and reduced flight speeds than those of agricultural aircraft, factors that contribute to improved target deposition and reduced lateral droplet displacement.

In addition, Zhang et al. (2023) demonstrated that spraying systems equipped with electronic Pulse Width Modulation (PWM) control enable dynamic adjustment of spray flow during operation, allowing precision variable-rate pesticide application according to crop characteristics. Consequently, continuous advances in technologies such as rotary atomizers and electronic spray control systems have directly improved application uniformity. The technologies incorporated into agricultural drones provide greater control over pesticide distribution, increasing deposition efficiency on the target while promoting a more rational and cost-effective use of agricultural inputs.

Pretto et al. (2019) also reported that new drone models with greater payload capacities and lower operating costs than conventional agricultural aircraft are becoming increasingly important. At the same time, the growing demand for sustainable agricultural practices is accelerating the adoption of more efficient and environmentally sustainable technologies, making spraying drones an increasingly viable alternative (Morales-Rodriguez et al., 2022).

According to IBGE (2017), approximately 77% of Brazilian farms occupy less than 50 ha. However, technology adoption depends not only on farm size but also on crop type. The state of Mato Grosso is dominated by extensive soybean, maize, and cotton production grown over large, contiguous areas with high operational demands, favoring the use of manned agricultural aircraft. In contrast, the state of São Paulo is characterized by perennial crops, particularly citrus, in addition to sugarcane, cultivated under more fragmented conditions and variable topography. Under these conditions, spraying drones offer greater operational flexibility, encouraging their adoption.

This study has some limitations that should be considered when interpreting the results. First, the analysis focused exclusively on equipment acquisition costs and did not include operating costs, such as fuel, labor, and routine maintenance, which can substantially influence long-term cost-effectiveness. In addition, operating multiple drones simultaneously requires a larger number of qualified operators and greater logistical support, resulting in additional costs that were not incorporated into the model. Future studies should include analyses of the Total Cost of Ownership (TCO), regional labor costs, and equipment life-cycle costs to provide a more comprehensive comparison of the technologies evaluated.

Conclusions

The mathematical model developed in this study enabled a quantitative assessment of the operational capacity of manned agricultural aircraft and spraying drones at different application rates, providing evidence of their operational efficiency and economic viability under different production scenarios. The main findings are as follows:

1. Operational equivalence: To achieve a daily application capacity of 900 ha day⁻¹ (the operational capacity of the EMB-203 at 5 L ha⁻¹), approximately five AGRAS T40, seven AGRAS T30, or ten AGRAS T20P drones are required, demonstrating that replacing a single manned aircraft requires the simultaneous operation of multiple unmanned units.

2. Cost-effectiveness: Among the drone models evaluated, the AGRAS T40 exhibited the best cost-effectiveness. The initial investment (one drone, three batteries, one charger, and one generator) was approximately 3.6 times lower than that of a new EMB-203 and 1.2 times lower than that of a used EMB-202 to achieve an equivalent daily application capacity.

3. Farm size scale: Spraying drones proved more suitable for properties of up to approximately 550 ha, a predominant farm-size profile in the state of São Paulo, whereas manned agricultural aircraft were better suited to large-scale operations, such as those found in Mato Grosso. These findings highlight the complementary roles of the two technologies according to farm size.

4. Geographic distribution: Kernel Density Estimation (KDE) identified a concentration of manned agricultural aircraft in the Northern and Southeastern mesoregions of Mato Grosso, whereas spraying drones were more concentrated in regions such as Ribeirão Preto, Triângulo Mineiro, and southern Rio Grande do Sul, with potential for expansion into agricultural areas of Northeast Brazil.

A limitation of this study is that the analysis did not include operating costs associated with fuel consumption, electricity, preventive and corrective maintenance, or specialized labor. Therefore, the results should be interpreted as a comparison based primarily on operational capacity and equipment acquisition costs.

Overall, the results indicate that spraying drones do not fully replace manned agricultural aircraft in terms of individual operational capacity. However, they represent an operationally and economically competitive alternative for small- and medium-sized farms or as complementary application platforms. These findings suggest that integrating manned agricultural aircraft and spraying drones offers a promising operational strategy for Brazilian agriculture by increasing technological flexibility and optimizing application capacity across different production scales.

References

• Althman, M. P. F. (2021). Avaliação da faixa de deposição em função do modelo de aeronave e volume de calda aplicado [Master's thesis, Universidade Estadual Paulista]. https://repositorio.unesp.br/server/api/core/bitstreams/051c27df-fc81-48f3-8837-73e03f288e1f/content
  » https://repositorio.unesp.br/server/api/core/bitstreams/051c27df-fc81-48f3-8837-73e03f288e1f/content
• Agência Nacional de Aviação Civil (ANAC) (2023). Regulamento Brasileiro da Aviação Civil nº 137 (RBAC nº 137): Operações aeroagrícolas. https://www.anac.gov.br
  » https://www.anac.gov.br
• Agência Nacional de Aviação Civil (ANAC) (2024). Consulta ao banco de dados do Registro Aeronáutico Brasileiro. https://www.gov.br/anac/pt-br/acesso-a-informacao/dados-abertos/areas-de-atuacao/aeronaves-1/drones-cadastrados
  » https://www.gov.br/anac/pt-br/acesso-a-informacao/dados-abertos/areas-de-atuacao/aeronaves-1/drones-cadastrados
• Antuniassi, U. R., & Boller, W. (2019). Tecnologia de aplicação culturas Anuais. Aldeia Norte.
• Baio, F. H. R., Antuniassi, U. R., Castilho, P. S., Teodoro, P. E., & Da Silva, E. E. (2019). Factors affecting aerial spray drift in the Brazilian Cerrado. PLOS ONE, 14(2), e0212289. https://doi.org/10.1371/journal.pone.0212289
  » https://doi.org/10.1371/journal.pone.0212289
• Brelje, B. J., & Martins, J. R. R. A. (2019). Electric, hybrid, and turboelectric fixed-wing aircraft: A Review of Concepts, Models, and Design Approaches. Progress in Aerospace Sciences, 104, 1-19. https://doi.org/10.1016/j.paerosci.2018.06.004
  » https://doi.org/10.1016/j.paerosci.2018.06.004
• Carvalho, F. K., Chechetto, R. G., Mota, A. A. B., & Antuniassi, U. R. (2021). Entendendo a tecnologia de aplicação, aviões, helicópteros e drones de pulverização. FEPAF.
• Carvalho, F. K., & Cunha, J. P. A. (2019). Atomizadores rotativos para aplicação aérea. In U. R. Antuniassi & W. Boller (Eds.), Tecnologia para culturas anuais. Aldeia Norte.
• Carvalho, F. F., & Oliveira, T. M. G. (2019). Defensivos agrícolas: aspectos técnicos, regulamentares e ambientais. Agronômica Ceres.
• Chen, P., Douzals, J. P., Lan, Y., Cotteux, E., Delpuech, X., Pouxviel, G., & Zhan, Y. (2022). Characteristics of unmanned aerial spraying systems and related spray drift: A review. Frontiers in Plant Science, 13, 870956. https://doi.org/10.3389/fpls.2022.870956
  » https://doi.org/10.3389/fpls.2022.870956
• Cunha, J. P. A. R., Barizon, R. R. M., Ferracini, V. L., Assalin, M. R., & Antuniassi, U. R. (2017). Spray drift and pest control from aerial applications on soybeans. Engenharia Agrícola, 37(3), 493-501. https://doi.org/10.1590/1809-4430-Eng.Agric.v37n3p493-501/2017
  » https://doi.org/10.1590/1809-4430-Eng.Agric.v37n3p493-501/2017
• De Miranda, E. E. (2018). Áreas cultivadas no Brasil e no mundo. AgroANALYSIS.
• Harris, C. R. , Millman, K. J. , van der Walt, S. J. , Gommers, R. , Virtanen, P. , Cournapeau, D. , Wieser, E. , Taylor, J. , Berg, S. , Smith, N. J. , Kern, R. , Picus, M. , Hoyer, S. , van Kerkwijk, M. H. , Brett, M. , Haldane, A. , del Río, J. F. , Wiebe, M. , Peterson, P. , et al. ( 2020 ). Array programming with NumPy. Nature , 585 , 357-362. https://doi.org/10.1038/s41586-020-2649-2
  » https://doi.org/10.1038/s41586-020-2649-2
• Hunter, J. D. (2007). Matplotlib: A 2D graphics environment. Computing in Science & Engineering, 9(3), 90-95. https://doi.org/10.1109/MCSE.2007.55
  » https://doi.org/10.1109/MCSE.2007.55
• Instituto Brasileiro de Geografia e Estatística (IBGE). (2017). Censo Agropecuário 2017, Resultados definitivos Brasil. https://censoagro2017.ibge.gov.br/resultados-censo-agro-2017/resultados-definitivos.html
  » https://censoagro2017.ibge.gov.br/resultados-censo-agro-2017/resultados-definitivos.html
• Instituto Nacional de Colonização e Reforma Agrária (INCRA/SIGEF). (2024). Sistema de Gestão Fundiária Brasileiro. https://sigef.incra.gov.br/
  » https://sigef.incra.gov.br/
• Kern, S. M. (2023). Relato técnico sobre geotecnologia e agricultura de precisão. ALLCOMP Geotecnologia e Agricultura.
• Lan, Y., Chen, S., & Fritz, B. K. (2017). Current status and future trends of precision agricultural aviation technologies. International Journal of Agricultural and Biological Engineering, 10(3), 1-17. https://doi.org/10.3965/j.ijabe.20171003.3088
  » https://doi.org/10.3965/j.ijabe.20171003.3088
• Luenberger, D. G. (2016). Investment science. Oxford University Press.
• Massey, M., O'Cass, A., & Otahal, P. A. (2018). Meta-analytic study of the factors driving the purchase of organic food. Appetite, 125, 418-427. https://doi.org/10.1016/j.appet.2018.02.029
  » https://doi.org/10.1016/j.appet.2018.02.029
• McKinney, W. (2010). Data structures for statistical computing in Python. In Proceedings of the 9th Python in Science Conference (pp. 56-61). https://doi.org/10.25080/Majora-92bf1922-00a
  » https://doi.org/10.25080/Majora-92bf1922-00a
• Melo, P. C. (2020). Análise de viabilidade econômica de projetos Editora Atlas.
• Nascimento, J. M., Gavassoni, W. L., Bacchi, L. M. A., Zuntini, B., Melo, E. P., & Leonel, R. K. (2018). Número de aplicações e uso de adjuvantes, adicionados a fungicidas no controle da ferrugem asiática da soja. Agrarian, 11(40), 95-104.
• Morales-Rodriguez, P. A., Cano, E. C., & Villena, J. A. (2022). Comparison between Conventional Sprayers and New UAV Sprayers: A Study Case of Vineyards and Olives in Extremadura (Spain). Agronomy, 12(6), 1307. https://doi.org/10.3390/agronomy12061307
  » https://doi.org/10.3390/agronomy12061307
• Mwitta, C., & Rains, G. C. (2024). The integration of GPS and visual navigation for autonomous navigation in agricultural fields. Frontiers in Robotics and AI, 11, 1359887. https://doi.org/10.3389/frobt.2024.1359887
  » https://doi.org/10.3389/frobt.2024.1359887
• Oliveira, L. H. A. (2021). Avanços na produção de defensivos agrícolas: Efeitos sobre o meio ambiente e a saúde humana. Agropecuária.
• Pacheco, E. P. (2000). Seleção e custo operacional de máquinas agrícolas. Embrapa Acre. https://ainfo.cnptia.embrapa.br/digital/bitstream/CPAF-AC/3959/1/doc58.pdf
  » https://ainfo.cnptia.embrapa.br/digital/bitstream/CPAF-AC/3959/1/doc58.pdf
• Pretto, A., Chebrolu, N., Burgard, W., & Stachniss, C. (2019). Building an Aerial-Ground Robotics System for Precision Farming: An Adaptable Solution. Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2019) Macau, China, 5951-5958. https://doi.org/10.1109/IROS40897.2019.8968116
  » https://doi.org/10.1109/IROS40897.2019.8968116
• Schimmelpfennig, D. (2016). Farm Profits and Adoption of Precision Agriculture. USDA Economic Research Service, Economic Research Report No. 217. https://www.ers.usda.gov/publications/80325
  » https://www.ers.usda.gov/publications/80325
• Silva, L. B. G. J., & Henkes, J. A. (2025). Uma análise da aviação agrícola no Brasil: Contexto e caracterização. Revista Brasileira de Aviação Civil & Ciências Aeronáuticas. https://rbac.cia.emnuvens.com.br/revista/article/view/372
  » https://rbac.cia.emnuvens.com.br/revista/article/view/372
• Souza, J. B. (2020). Gestão da agricultura de precisão: Otimização de processos e resultados. Editora Agrícola.
• Sistema Integrado de Comércio Exterior (SISCOMEX). (2024). Sistema Integrado de Comércio Exterior. Receita Federal do Brasil. https://www.gov.br/receitafederal/pt-br/assuntos/aduana-e-comercio-exterior/importacao-e-exportacao/sistema-integrado-de-comercio-exterior-siscomex
  » https://www.gov.br/receitafederal/pt-br/assuntos/aduana-e-comercio-exterior/importacao-e-exportacao/sistema-integrado-de-comercio-exterior-siscomex
• Van Rossum, G., & Drake, F. L. (2009). Python 3 reference manual CreateSpace.
• Wang, L., Lan, Y., Zhang, Y., Zhang, H., Tahir, M. N., Ou, S., Liu, X., & Chen, P. (2019). Applications and Prospects of Agricultural Unmanned Aerial Vehicle Obstacle Avoidance Technology in China. Sensors, 19(3), 642. https://doi.org/10.3390/s19030642
  » https://doi.org/10.3390/s19030642
• Wang, Q., & Zhang, S. (2021). Applying drone-based spatial mapping to help growers manage crop diseases. Journal of Extension, 59(2), Article 11. https://doi.org/10.34068/joe.59.02.11
  » https://doi.org/10.34068/joe.59.02.11
• Waskom, M. L. (2021). seaborn: Statistical data visualization. Journal of Open Source Software, 6(60), 3021. https://doi.org/10.21105/joss.03021
  » https://doi.org/10.21105/joss.03021
• Zhang, Z., Yuan, S., Lai, Q., Zeng, R., Zhang, J., & Shen, S. (2023). Precision variable-rate control system for mini-UAV-based pesticide application. Journal of Physics: Conference Series, 2557(1), 012006. https://doi.org/10.1088/1742-6596/2557/1/012006
  » https://doi.org/10.1088/1742-6596/2557/1/012006

Data Availability Statement

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

* Corresponding author:

[e-mail] | [e-mail]

Area Editor:

Murilo Mesquita Baesso

Publication Dates

• Publication in this collection
  26 Oct 2026
• Date of issue
  2026

History

• Received
  1 Apr 2025
• Accepted
  10 June 2026