Search

Site-specific application of herbicides is highly desirable for optimising its usage and reducing environmental damages. Thus, developing techniques for identification and mapping of weeds is necessary for a proper precision agriculture adoption. Such weed identification for site-specific management is difficult when the main crop is already established in the field. This study shows the possibility of differentiating sugarcane plants from weeds by the spectral behaviour of the leaves. The performance of two modelling methods, SIMCA (soft independent modelling by class analogy) and the RF algorithm (random forest) was compared. The simplification of the Vis-NIR spectrum into only four bands of interest (500–550 nm; 650–750 nm; 1300–1450 nm; and 1800–1900 nm) was verified by demonstrating they had the same differentiation ability as the full visible-near infra-red spectrum. Thus, it was shown that performing the proper band selection and local calibration using a spectral classification approach may allow weed mapping and facilitate localised herbicide application.

Data interpolation is widely required in precision agriculture. However, its effectiveness is a function of the characteristics of the dataset, being necessary for the evaluation of several parameters. This study aimed to identify how the spatial interpolators, Kriging, and Inverse Distance Weighting, are influenced by the degree of spatial dependence of the variables analyzed and the number of neighbors considered in the interpolation process (sampling neighborhood). Soil samples were collected from three sugarcane fields. By the optimization process, we verified that the sampling neighborhood influences the accuracy of interpolations, but there is not a standard recommendation to follow. Thus, the best sampling neighborhood must ever be optimized for each case when preparing fertilizer prescription maps. Evaluating the performance of interpolations is always important to infer the reliability of the prescription maps, since no index that measures the degree of spatial dependence is effective. Because high prediction errors can occur when spatial dependence is poorly modeled, one cannot expect crop response with the continuous application of fertilizers in variable rates. Thus, work with homogeneous soil zones can be an interesting palliative approach. This study guides precision agriculture practitioners on some points that should be carefully considered in the data interpolation process for generating fertilizer prescription maps.

Sampling density is a parameter that allows to identify the spatial variability of properties of interest in soils and is essential for precision agriculture practices, such as fertilizer application in variable rate. Soil sensors and spectroscopic techniques have been investigated and reported as promising tools, although showing some limitation in characterizing the chemical fertility of soils. One of the causes of the limitations is the way samples are prepared. Therefore, the aim of this work was to evaluate the effect of sieving and of drying temperature, on the determination of total nitrogen and phosphorus and potassium contents, by using Vis-NIR and Mid-Infrared spectroscopies. We found that Diffuse Reflectance Spectroscopy in the Visible and Near-Infrared region of the spectra shows better performance of the prediction models than of the Attenuated Total Reflectance Spectroscopy in the Mid-Infrared region. We did not identify significant influence of drying temperature neither of soil particle size on the predictive quality of the models generated via Vis-NIR spectroscopy; thus, the extra expenditure of time and cost for a more intense preparation of the soil samples would not be justifiable. In addition, we infer that the high prediction error may limit the applicability of spectroscopy for guiding the variable-rate application of fertilizers within the scope of precision agriculture.

Precision agriculture aspires to manage the soil spatial variability within a field. However, agricultural practices can decrease the spatial autocorrelation of soil properties at a short distance, which requires an increase in sampling density to improve the reliability of interpolated surfaces. Data fusion of several soil spectrometers is a promising alternative that can increase knowledge about soil physicochemical variability. Therefore, this research aimed to evaluate whether the synergy of portable X-ray fluorescence, Vis-NIR, and Mid-Infrared spectrometers can successfully increase sampling density and improve the quality of the soil attribute maps, in a high fertilization management study area. We conducted a case study on a sugarcane field in Southeast Brazil. Our spectral data fusion approach did not increase the prediction performance for soil properties, achieving the same accuracy (α ≤ 5%) of the single spectral ranges (Vis-NIR or MIR or pXRF) individually. Thus, Vis-NIR was chosen to evaluate if spectroscopy could improve soil properties prediction and mapping. We used three dataset scenarios to build soil property maps: two were based on soil wet-chemical laboratory analysis using 1 sample per ha, and 1 sample per 0.25 ha and a third used a spectroscopy model with 0.25 samples per ha calibrated using soil wet-chemical laboratory analysis at the rate of 1 sample per ha. In a field with strong anthropic influences, due to liming and soil fertilization, Vis-NIR spectroscopy cannot predict well the amount of phosphorus and pH at a sampling density of 1 sample per 0.25 ha. Increased soil sampling enhances soil mapping accuracy only for chemical properties (e.g., potassium) that are correlated with soil texture. Nevertheless, soil spectra can classify areas of similar soil texture and help farmers with the site-specific management of crops.

Nitrogen (N) fertilization is challenging for sugarcane, and machine-based canopy sensors appear as an alternative to allow variable-rate N fertilization. Top or sidedressing N is applied in each crop row and crop spatial variability behavior must be understood to allow proper sensor placement and applicator configurations in order to optimize N fertilization. Thus, the goal of this study was to investigate sugarcane crop variability and N prescription error when working with various sensor placements and boom sections. The approaches involved post-processing N prescription maps and real-time application, varying the number of sensors used and calculating the N rate for the applicator boom sections. Sugarcane fields show high crop variability due to their semi-perennial cropping system, which causes unpredictability of sensor readings from adjacent rows, ideally suggesting one sensor for each row in order to obtain more detailed plant-vigor information. Moreover, the machine must be able to apply fertilizer for each individual row to allow the most reliable application of N rate, ensuring optimization of crop response to variable-rate N application.

O objetivo deste trabalho foi identificar as diferenças entre o banco de sementes de área cultivada com cana-de-açúcar colhida mecanicamente e o de área colhida após a queima do canavial, utilizando-se técnicas de agricultura de precisão. As amostragens de solo para determinação do banco de sementes foram realizadas utilizando-se grade amostral regular de 2 ha. Os mapas de infestação foram obtidos pela técnica de interpolação por "krigagem". Observou-se que o talhão de cana crua possui menor potencial de infestação de plantas daninhas em relação ao talhão de cana queimada, principalmente monocotiledôneas; entretanto, algumas dicotiledôneas podem ser selecionadas, como as do gênero Ipomoea. Conclui-se que a agricultura de precisão pode ser uma ferramenta útil para determinar mapas de infestação de plantas daninhas e que a palha de cana-de-açúcar em sistemas de cana crua pode ser utilizada como fator de supressão de várias espécies.

The objective of this work was to evaluate whether a canopy sensor is capable of estimating sugarcane response to N, as well as to propose strategies for handling the data generated by this device during the decision-making process for crop N fertilization. Four N rate-response experiments were carried out, with N rates varying from 0 to 240 kg ha-1. Two evaluations with the canopy sensor were performed when the plants reached average stalk height of 0.3 and 0.5 m. Only two experiments showed stalk yield response to N rates. The canopy sensor was able to identify the crop response to different N rates and the relationship of the nutrient with sugarcane yield. The response index values obtained from the canopy sensor readings were useful in assessing sugarcane response to the applied N rate. Canopy reflectance sensors can help to identify areas responsive to N fertilization and, therefore, improve sugarcane fertilizer management. Resumo:O objetivo deste trabalho foi avaliar se um sensor de dossel é capaz de estimar a resposta da cana-de-açúcar ao N, bem como propor estratégias para tratar os dados gerados pelo equipamento durante o processo de tomada de decisão para a adubação nitrogenada da cultura. Foram conduzidos quatro experimentos de dose-resposta ao N, com doses de N que variaram de 0 a 240 kg ha-1. Duas avaliações foram realizadas com o sensor de dossel quando as plantas alcançaram altura média de colmos de 0,3 e 0,5 m. Apenas dois experimentos mostraram resposta ao N aplicado em termos de produtividade de colmos. O sensor foi capaz de identificar a resposta da cultura a diferentes doses de N e a relação do nutriente com a produtividade da cana. Os índices de resposta obtidos pelas leituras do sensor foram úteis em avaliar a resposta da cana-de-açúcar ao N aplicado. Sensores de dossel podem auxiliar na identificação de áreas responsivas à aplicação de N e, consequentemente, melhorar o manejo da fertilização nitrogenada da cana-de-açúcar.

Precision agriculture is crucial for increasing food output without expanding the cultivable area, which requires sensors to be deployed for controlling the level of nutrients in the soil. In this paper, we report on a microfluidic electronic tongue (e-tongue) based on impedance measurements which is capable of distinguishing soil samples enriched with plant macronutrients. The e-tongue setup consisted of an array of sensing units made with layer-by-layer films deposited onto gold interdigitated electrodes. Significantly, the sensing units could be reused with adequate reproducibility after a simple washing procedure, thus indicating that there is no cross-contamination in three independent sets of measurements. A high performance was achieved by treating the capacitance data with the multidimensional projection techniques Principal Component Analysis (PCA), Interactive Document Map (IDMAP), and Sammon’s Mapping. While an optimized performance was demonstrated with IDMAP and feature selection, during which data of a limited frequency range were used, the distinction of all soil samples was also possible with the well-established PCA analysis for measurements at a single frequency. The successful use of a simple microfluidic e-tongue for soil analysis paves the way for enhanced tools to support precision agriculture.

Nowadays, one of the biggest issues addressed to electronic sensor fabrication is the build-up of efficient electrodes as an alternative way to the expensive, complex and multistage processes required by traditional techniques. Printed electronics arises as an interesting alternative to fulfill this task due to the simplicity and speed to stamp electrodes on various surfaces. Within this context, the Fused Deposition Modeling 3D printing is an emerging, cost-effective and alternative technology to fabricate complex structures that potentiates several fields with more creative ideas and new materials for a rapid prototyping of devices. We show here the fabrication of interdigitated electrodes using a standard home-made CoreXY 3D printer using transparent and graphene-based PLA filaments. Macro 3D printed electrodes were easily assembled within 6 min with outstanding reproducibility. The electrodes were also functionalized with different nanostructured thin films via dip-coating Layer-by-Layer technique to develop a 3D printed e-tongue setup. As a proof of concept, the printed e-tongue was applied to soil analysis. A control soil sample was enriched with several macro-nutrients to the plants (N, P, K, S, Mg, and Ca) and the discrimination was done by electrical impedance spectroscopy of water solution of the soil samples. The data was analyzed by Principal Component Analysis and the 3D printed sensor distinguished clearly all enriched samples despite the complexity of the soil chemical composition. The 3D printed e-tongue successfully used in soil analysis encourages further investments in developing new sensory tools for precision agriculture and other fields exploiting the simplicity and flexibility offered by the 3D printing techniques.