How to perform precision seeding

Precision agriculture tools help identify and plan the most appropriate scenarios and routes with the aim of making the seeding operation more precise.

Several crop management techniques have been made available in recent years, due to the change in the profile of Brazilian agriculture, causing producers to seek techniques that increase the productive capacity of their properties, while remaining competitive in the market. The set of these techniques, called Precision Agriculture (AP), has as its principle the use of available resources in a rational and precise way combined with technology systems, in order to reduce the cost and increase the profitability of agricultural activity.

When thinking about cost reduction, it is essential that operations are carried out with as little expenditure as possible. The success or failure of the crop is related to the sowing operation, considered one of the most important stages of the production process. When this operation is successful, it facilitates crop management, improving the efficiency of subsequent operations, mainly the application of pesticides and mechanized harvesting.

Read also

Thus, the use of software, combined with an automatic targeting system, can greatly contribute to the implementation and profitability of an agricultural system, as it allows planning the area and estimating expenses in the sowing operation. Furthermore, it enables comparative analyzes of possible routes to be followed by the tractor-seeder set, helping to choose the most suitable one to be carried out in the field, based on a series of requirements, such as number of maneuvers, length of lines, in addition to the estimation of operational capacity. It is worth mentioning that the application and effectiveness of these software in the field are only possible thanks to Global Navigation Satellite Systems (GNSS).

Among the GNSS positioning methods, the most used today is relative positioning using an RTK signal (Real time kinematic), which guarantees maximum errors of around 2,5cm. However, it is worth remembering that this type of positioning, when used in large areas, or in areas with uneven terrain, may present signal degradation, making it necessary to install repeater bases to ensure signal quality.

Planning operations using a project and using a GNSS signal, regardless of the type of signal, improves the quality of service, since the operator has to divert his attention to directing the machine.

Thus, the use of this software, such as AgroCAD, has helped producers in several aspects, such as choosing the best route for the machines, so that the farmer can choose the route in which the machine will spend more time in operation and less time making maneuvers. , in addition to estimating, adjusting and systematizing the area for subsequent operations. Therefore, to evaluate the effectiveness of the AgroCAD software, the Agricultural Machinery and Mechanization Laboratory (Lamma) carried out an experiment simulating scenarios for sowing peanuts in the region of Orindiúva (SP), in which three scenarios were generated for the operation, selecting The most appropriate scenario for carrying out the operation in the field was then compared, then the scenario executed was compared with the one previously designed in the software.

To this end, the AgroCAD software, developed by TecGraf, an Autodesk representative, was used to plan the sowing lines.

In the field, the boundaries of the plots were surveyed, using the tractor's own system to record the route (AMS RTK John Deere). Then, the terraces were explored to survey their coordinates. With the fieldwork completed, using a pen drive, the files were transferred from the GS3 monitor to the AgroCAD software.

After the files were transferred, from the terraces mapped in the field, parallel lines were created with a spacing of 5,4 meters, compatible with the seeder configuration (six lines with a spacing of 0,9m). At Figure 1 shows the illustration of the parallel lines created, as well as the positioning of the sowing rows.

Once the lines were drawn, an analysis was carried out with the objective of verifying the best way to distribute the rows within the plot, carrying out simulations that allowed evaluating which would be the most efficient sowing project, using, as a selection criterion, the analysis of dead contour lines (“killing”), better operational field capacity and maneuver optimization.

In this way, three scenarios were created with the lines planned differently between them. These scenarios were analyzed with a view to subsequently selecting one for execution in the field. The analyzes were carried out in the software itself.

For each scenario, the software allows two planning conditions to be compared: without and with optimization. The “without optimization” condition represents the use of planned lines, however, without joining nearby lines. In the “with optimization” condition, the software checks the possibilities of joining nearby planned lines, using as criteria the maximum distance between nodes and the maximum angle allowed by the autopilot. For the analysis of this work, values ​​of 25 meters of distance between points (nodes) and 10º for maximum angulation were considered.

SCENARIOS OBTAINED IN THE SOFTWARE

Comparing the number of lines (Figure 2) without optimization with the number of lines after optimization, it is observed that there was a reduction in shorter lines (0m to 200m), of the order of 7,2%, 5,4 % and 15,7% for scenarios 1, 2 and 3, respectively, while the number of lines longer than 800m increased by 7,3%, 4,9% and .2% for scenarios 1, 2 and 3, respectively (Figure 2).

The analysis of the scenarios, for the variable number of lines, points to the use of optimized scenario 1 or 3, as they present a greater reduction in lines with a length of up to 400m (22,5% and 25,7%, respectively) and also a greater increase in longer lines (7,3% and 7,2%), compared to scenario 2.

Although the percentages of reduction of smaller lines and increase of larger lines are close, it appears that in scenario 1, the total number of lines is 330, while in scenario 3 it is 443. The greater the number of lines In the same area, a lower average line length is expected, an undesirable characteristic in agricultural operations, as it would result in a greater number of maneuvers. However, on the other hand, it can mean better use of the useful area.

When comparing the total distances covered (Figure 3), after optimization, for lines longer than 800 meters in length, there is an increase of around 4,7%, 5,1% and 7,3% for scenarios 1, 2 and 3, respectively. Likewise, these scenarios present a reduction in the total distance traveled of 10,1%, 11,7% and 14,6%, respectively, for lines with lengths less than 800 meters. This result was to be expected, since to optimize the lines, the software joins lines that are close together, therefore, the distance traveled for shorter lines tends to decrease and those of longer length tend to increase.

The number of maneuvers to be carried out (Figure 4) showed, in all scenarios, a reduction after the optimization of the sowing lines. This reduction was 5,5%, 5,4% and 9,4% for scenarios 1, 2 and 3, respectively, and is directly related to the shape of the plot. The best option is for the plots to be multiple widths of the width of the machines to be used, avoiding underutilization of the machines. It is then possible to reduce the number of maneuvers in the field, favoring route efficiency and, consequently, the field efficiency of the machines. agricultural operations.

Analyzing the time spent on maneuvers (Figure 5), as expected, it is observed that there was a reduction in time after optimizing the sowing lines. In scenarios 1, 2 and 3 this reduction was 5,2%, 5,4% and 9,5%, respectively. Regarding travel time, the three scenarios did not show considerable differences after optimizing the sowing lines.

Regarding the total execution time, as expected, there was a reduction after the optimization of the sowing lines, with a reduction of 1,3% for scenario 1; 1,4% for scenario 2 and 2,7% for scenario 3. This reduction in the time spent on maneuvers and the total execution time, after optimization, was already expected, since the software promotes the joining of lines that are close together, and, consequently, results in fewer maneuvers and an increase in the operational capacity of the mechanized set.

It is worth mentioning that to carry out a certain mechanized operation, time losses will always occur, which can be controlled and are likely to be reduced through rationalization and control of the activity carried out, or otherwise uncontrollable — influenced by climatic, personal factors and mechanized systems . Therefore, it is essential to have a study of times and movements, since this is a basic instrument for optimizing mechanized operations, obtaining better execution and greater savings with the subdivision of all movements carried out during the operation.

When it comes to route efficiency, still in Figure 5, after optimization, scenario 2 presents the highest efficiency (74%), followed by scenarios 1 and 3, which present efficiencies of 73% and 70%, respectively. However, it appears that the difference in route efficiency was minimal between the three scenarios analyzed and can be considered adequate for the peanut sowing operation.

PEANUT SOWING

After analyzing all variables, optimized scenario 1 was selected for execution, as it presents a smaller total number of lines, a shorter total distance covered, a smaller number of maneuvers and, consequently, less time spent on maneuvers and a shorter total time. execution.

Still using the AgroCAD software, the selected scenario was transformed into John Deere format, and then saved on a pen drive so that it could be transferred to the GS3 monitor.

In the John Deere format, unlike other brands, the files can only be interpreted by the John Deere system and, therefore, it does not allow ready lines in “shapefile” (universal file format) to be inserted on the monitor, requiring a conversion through software.

After inserting the planned lines into the monitor, the tractor operator is only responsible for carrying out the maneuvers at the end of each planned line, aligning the tractor in the subsequent line and activating the autopilot to carry out the seeding.

Sowing was carried out using a John Deere 7815 tractor, which has 148kW (202hp) of engine power, 4x2 TDA traction, operating at an average travel speed of 7km/h, equipped with a Star Fire GPS receiver, model SF3000; Green Star monitor, model 3 2630, with RTK correction system with 900MHz radio and hydraulic autopilot. A Marchesan precision seeder-fertilizer, Cop Suprema model, was used, with six sowing lines, equipped with a 15” (38,1cm) offset double disc for seed deposition and a pneumatic distributor. The spacing adopted between rows was 0,90m with a sowing density of around 20 seeds per meter.

COMPARISONS BETWEEN THE SCENARIOS

After executing the selected scenario (end of sowing), data was extracted from the lines executed and, from there, a comparison was made between the selected scenario and the scenario executed in the field. Using AgroCAD software commands, reliability analyzes of the project execution were carried out, which made it possible to identify deviations made in the field, together with the comparison of maneuver time, operational field capacity and effective sowing time.

It is observed that there was an increase of 5,7% and 4,6% for the length of lines from zero to 200 meters and 200 to 400 meters, respectively. While for lines longer than 800 meters it reduced by 1% (Figure 6A), and the total distance traveled reduced by 0,6% for lines with lengths less than 800 meters and increased by 0,1% for lines longer than 800 meters, of the selected scenario with that of the executed scenario (Figure 6B).

In Figure 7, it is possible to notice that there was a 6,8% reduction in the average length of the lines and, in addition, in the same proportion the increase in the number of maneuvers from those selected to those executed in the field.

Considering that the number of maneuvers increased, consequently the time spent on maneuvers also increased by 6,8% (Figure 8). On the other hand, the travel time did not show a considerable difference, while the total execution time increased by around 1,9% from planned to executed, however, maintaining constant route efficiency throughout the operation (73%).

The differences found between the scenarios (selected and executed) for maneuver and total execution times are small (6,8% and 1,9% higher, respectively) and are justified, as are the differences found when comparing the other variables. , due to the fact that, during execution, influences from adverse factors may occur. These factors can be, among others, the presence of obstacles and failures in labor and systematization of the terrain. These adversities are not added to the AgroCAD software and, therefore, can lead to an increase in variables in the execution of the peanut sowing operation.

It should also be noted that, despite these small differences, the use of AgroCAD software was beneficial, as it allows the selection of a scenario that favors the efficiency of the sowing operation, to the detriment of other options that would make the process more onerous, since there would be more time spent on maneuvers and a greater number of short lines. Furthermore, the software is efficient in carrying out scenario analyzes and enables their optimization, proving to be a good tool to assist the farmer in carrying out sowing. 

The use of the software made it possible to optimize peanut sowing lines, increasing the average length of the lines and consequently increasing the number of shots from the mechanized set at the time of sowing. With the planning of sowing lines in the software, it is possible to align subsequent operations, in order to reduce costs with agricultural activity, as well as establish traffic control in the planned areas, since the machines will always travel under the same lane, regardless of the operation, it will only depend on the gauge adjustment of these.

Rouverson Pereira da Silva, Gabriel Garcia Blumer, Adão Felipe dos Santos, Luiz Augusto de S. Nardo, Cristiano Zerbato, Carlos Eduardo Angeli Furlani, Agricultural Machinery and Mechanization Laboratory – Lamma, Unesp-Jaboticabal, SP

Article published in issue 168 of Cultivar Máquinas