In soybean, there is a lack of research aiming to compare the performance of machine learning (ML) and deep learning (DL) methods to predict more than one agronomic variable, such as days to maturity (DM), plant height (PH), and grain yield (GY). As these variables are important to developing an overall precision farming model, we propose a machine learning approach to predict DM, PH, and GY for soybean cultivars based on multispectral bands. The field experiment considered 524 genotypes of soybeans in the 2017/2018 and 2018/2019 growing seasons and a multitemporal–multispectral dataset collected by embedded sensor in an unmanned aerial vehicle (UAV). We proposed a multilayer deep learning regression network, trained during 2000 epochs using an adaptive subgradient method, a random Gaussian initialization, and a 50% dropout in the first hidden layer for regularization. Three different scenarios, including only spectral bands, only vegetation indices, and spectral bands plus vegetation indices, were adopted to infer each variable (PH, DM, and GY). The DL model performance was compared against shallow learning methods such as random forest (RF), support vector machine (SVM), and linear regression (LR). The results indicate that our approach has the potential to predict soybean-related variables using multispectral bands only. Both DL and RF models presented a strong (r surpassing 0.77) prediction capacity for the PH variable, regardless of the adopted input variables group. Our results demonstrated that the DL model (r = 0.66) was superior to predict DM when the input variable was the spectral bands. For GY, all machine learning models evaluated presented similar performance (r ranging from 0.42 to 0.44) for each tested scenario. In conclusion, this study demonstrated an efficient approach to a computational solution capable of predicting multiple important soybean crop variables based on remote sensing data. Future research could benefit from the information presented here and be implemented in subsequent processes related to soybean cultivars or other types of agronomic crops.

Robotic technologies are affording opportunities to revolutionize the production of specialty crops (fruits, vegetables, tree nuts, and horticulture). They offer the potential to automate tasks and save inputs such as labor, fertilizer, and pesticides. Specialty crops are well known for their high economic value and nutritional benefits, making their production particularly impactful. While previous review papers have discussed the evolution of agricultural robots in a general agricultural context, this review uniquely focuses on their application to specialty crops, a rapidly expanding area. Therefore, we aimed to develop a state-of-the-art review to scientifically contribute to the understanding of the following: (i) the primary areas of robots’ application for specialty crops; (ii) the specific benefits they offer; (iii) their current limitations; and (iv) opportunities for future investigation. We formulated a comprehensive search strategy, leveraging Scopus^® and Web of Science™ as databases and selecting “robot” and “specialty crops” as the main keywords. To follow a critical screening process, only peer-reviewed research papers were considered, resulting in the inclusion of 907 papers covering the period from 1988 to 2024. Each paper was thoroughly evaluated based on its title, abstract, keywords, methods, conclusions, and declarations. Our analysis revealed that interest in agricultural robots for specialty crops has significantly increased over the past decade, mainly driven by technological advancements in computer vision and recognition systems. Harvesting robots have arisen as the primary focus. Robots for spraying, pruning, weed control, pollination, transplanting, and fertilizing are emerging subjects to be addressed in further research and development (R&D) strategies. Ultimately, our findings serve to reveal the dynamics of agricultural robots in the world of specialty crops while supporting suitable practices for more sustainable and resilient agriculture, indicating a new era of innovation and efficiency in agriculture.

The assessment of blueberry fruit quality is traditionally conducted through laboratory equipment. Despite its high accuracy, this method remains destructive, labor-intensive, time-consuming, and costly. Consequently, there is a pressing need for innovative solutions such as image-based and artificial intelligence (AI)-driven analysis. To address these limitations, this study aimed to analyze whether an approach based on mobile image-based analysis combined with machine learning (ML) algorithms could develop a non-destructive framework for evaluating blueberry fruit quality, specifically focusing on total soluble solids (TSS) and firmness. Firstly, we collected numerous blueberry samples during the maturity stage to construct a comprehensive dataset. These samples were meticulously analyzed in a laboratory for diameter, TSS, firmness, and color. Simultaneously, RGB images were captured using a mobile device. These images were processed to extract spectral bands (red, green, and blue). Eight ML algorithms were employed to develop predictive models capable of predicting the qualitative parameters of the blueberries. Initially, correlation analysis demonstrated that RGB images suggestively contribute to fruit quality assessment (r < 0.41). However, the integration of ML algorithms significantly enhanced the predictive accuracy of these models (R² = 0.71–0.99, MAE = 0.003–0.28, and RMSE = 0.004–0.31). This innovative approach differs from existing methods by providing a non-destructive, cost-effective, and efficient alternative to assess blueberry fruit quality. Our findings, therefore, confirm the applicability of mobile images for supporting high-quality blueberry harvesting. These advancements can potentially reduce labor costs, increase yield quality, and support advancements in precision agriculture.

The objectives of this work are (i) to classify soybean cultivars under different irrigation managements using hyperspectral data, looking for the best machine-learning algorithm for the classification and the input that improves the performance of the models. The experiment was implemented in the 2023/24 harvest in the experimental area of the Federal University of Mato Grosso do Sul, Câmpus Chapadão do Sul, Mato Grosso do Sul, and it was conducted in a strip scheme with seven cultivars subjected to irrigated and rainfed management. Sixty days after crop emergence, three leaves per plot were collected for evaluation by the hyperspectral sensor. The spectral data was then separated into 28 bands to reduce dimensionality. In this way, two databases were generated: one with all the spectral information provided by the sensor (WL) and one with the 28 spectral bands (SB). Each database was subjected to different machine-learning models to ascertain the improved accuracy of the models in distinguishing the different eucalyptus species. The models tested were artificial neural networks (ANN), decision trees (DT), linear regression (LR), M5P algorithm, random forest (RF), and support vector machine (SVM). The results demonstrate the effectiveness of machine-learning models in differentiating soybean management under rainfed and irrigated conditions, highlighting the advantage of hyperspectral data (WL) over selected spectral bands (SB). Models such as the support vector machine (SVM) showed the best levels of accuracy when using the entire available spectrum. On the other hand, artificial neural networks (ANN) performed well with spectral band data, demonstrating their ability to work with smaller data sets without compromising the classification.

The application of hyperspectral data in machine learning models can contribute to the rapid and accurate determination of caffeine content in coffee beans. This study aimed to identify the machine learning algorithm with the best performance for predicting caffeine content and to find input data for these models that can improve the accuracy of these algorithms. The coffee beans were harvested one year after the seedlings were planted. The fresh beans were taken to the spectroscopy laboratory (Laspec) at the Federal University of Mato Grosso do Sul, Chapadão do Sul campus, for spectral evaluation using a spectroradiometer. For the analysis, the dried coffee beans were ground and sieved for the quantification of caffeine, which was carried out using a liquid chromatograph on the Waters Acquity 1100 series UPLC system, with an automatic sample injector. The spectral data of the beans, as well as the spectral data of the roasted and ground coffee, were analyzed using machine learning (ML) algorithms to predict caffeine content. Four databases were used as input: the spectral information of the bean (CG), the spectral information of the bean with additional clone information (CG+C), the spectral information of the bean after roasting and grinding (CGRG) and the spectral information of the bean after roasting and grinding with additional clone information (CGRG+C). The caffeine content was used as an output to be predicted. Each database was subjected to different machine learning models: artificial neural networks (ANNs), decision tree (DT), linear regression (LR), M5P, and random forest (RF) algorithms. Pearson’s correlation coefficient, mean absolute error, and root mean square error were tested as model accuracy metrics. The support vector machine algorithm showed the best accuracy in predicting caffeine content when using hyperspectral data from roasted and ground coffee beans. This performance was significantly improved when clone information was included, allowing for an even more accurate analysis.

Robots are increasingly emerging as effective platforms to overcome a wide range of challenges in agriculture. Beyond functioning as standalone systems, agricultural robots are proving valuable as collaborative platforms, capable of supporting and integrating with humans and other technologies and agricultural activities. In this study, we designed and implemented an automated system embedded in a ground-based robotic platform to support spraying drone operations. The system consists of a robotic platform that carries the spraying drone along with all necessary support devices, including a water tank, chemical reservoirs, a mixer, generators for drone battery charging, and a top landing pad. The system is controlled with a mobile app that calculates the total amount of water and chemicals required and sends commands to the platform to prepare the application mixture. The input information in the app includes the field area, application rate, and up to three chemical dosages simultaneously. Additionally, the platform allows the drone to take off from and land on it, enhancing both safety and operability. A set of pumps was used to deliver water and chemicals as specified in the mobile app. To automate pump control, we used Arduino technology, including both the microcontroller and a programming environment for coding and designing the mobile app. To validate the system’s effectiveness, we individually measured the amount of water and chemical delivered to the mixer tank and compared it with conventional manual methods for calculating chemical quantities and preparation time. The system demonstrated consistent results, achieving high precision and accuracy in delivering the correct amount. This study advances the field of agricultural robotics by highlighting the role of collaborative platforms. Particularly, the system presents a valuable and low-cost solution for small farms and experimental research.