Machine Learning in Agriculture: A Comprehensive Updated Review

Predicting protein phosphorylation sites in soybean using interpretable deep tabular learning network

Phosphorylation of proteins is one of the most significant post-translational modifications (PTMs) and plays a crucial role in plant functionality due to its impact on signaling, gene expression, enzyme kinetics, protein stability and interactions. Accurate prediction of plant phosphorylation sites (p-sites) is vital as abnormal regulation of phosphorylation usually leads to plant diseases. However, current experimental methods for PTM prediction suffers from high-computational cost and are error-prone. The present study develops machine learning-based prediction techniques, including a high-performance interpretable deep tabular learning network (TabNet) to improve the prediction of protein p-sites in soybean. Moreover, we use a hybrid feature set of sequential-based features, physicochemical properties and position-specific scoring matrices to predict serine (Ser/S), threonine (Thr/T) and tyrosine (Tyr/Y) p-sites in soybean for the first time. The experimentally verified p-sites data of soybean proteins are collected from the eukaryotic phosphorylation sites database and database post-translational modification. We then remove the redundant set of positive and negative samples by dropping protein sequences with >40% similarity. It is found that the developed techniques perform >70% in terms of accuracy. The results demonstrate that the TabNet model is the best performing classifier using hybrid features and with window size of 13, resulted in 78.96 and 77.24% sensitivity and specificity, respectively. The results indicate that the TabNet method has advantages in terms of high-performance and interpretability. The proposed technique can automatically analyze the data without any measurement errors and any human intervention. Furthermore, it can be used to predict putative protein p-sites in plants effectively. The collected dataset and source code are publicly deposited at https://github.com/Elham-khalili/Soybean-P-sites-Prediction.

Modeling Reference Crop Evapotranspiration Using Support Vector Machine (SVM) and Extreme Learning Machine (ELM) in Region IV-A, Philippines

The need for accurate estimates of reference crop evapotranspiration (ETo) is important in irrigation planning and design, irrigation scheduling, reservoir management among other applications. ETo can be accurately determined using the internationally accepted FAO Penman–Monteith (FAO-56 PM) equation. However, this requires numerous observed data, including solar radiation, air temperature, relative humidity, and wind speed, which in most cases are unavailable, particularly in developing countries such as the Philippines. This study developed models based on Support Vector Machines (SVMs) and Extreme Learning Machines (ELMs) for the estimation of daily ETo using different input combinations of meteorological data in Region IV-A, Philippines. The performance of machine learning models was compared with the different established alternative empirical models for ETo. The results show that the SVM and ELM models, with at least Tmax, Tmin, and Rs as inputs, provide the best daily ETo estimates. The accuracy of machine learning models was also found to be superior compared to the empirical models given with same input requirements. In general, SVM and ELM models showed similar modeling performance, although the former showed lower run time than the latter

Combining Spectral and Textural Information from UAV RGB Images for Leaf Area Index Monitoring in Kiwifruit Orchard

The use of a fast and accurate unmanned aerial vehicle (UAV) digital camera platform to estimate leaf area index (LAI) of kiwifruit orchard is of great significance for growth, yield estimation, and field management. LAI, as an ideal parameter for estimating vegetation growth, plays a significant role in reflecting crop physiological process and ecosystem function. At present, LAI estimation mainly focuses on winter wheat, corn, soybean, and other food crops; in addition, LAI on forest research is also predominant, but there are few studies on the application of orchards such as kiwifruit. Concerning this study, high-resolution UAV images of three growth stages of kiwifruit orchard were acquired from May to July 2021. The extracted significantly correlated spectral and textural parameters were used to construct univariate and multivariate regression models with LAI measured for corresponding growth stages. The optimal model was selected for LAI estimation and mapping by comparing the stepwise regression (SWR) and random forest regression (RFR). Results showed the model combining texture features was superior to that only based on spectral indices for the prediction accuracy of the modeling set, with the R2 of 0.947 and 0.765, RMSE of 0.048 and 0.102, and nRMSE of 7.99% and 16.81%, respectively. Moreover, the RFR model (R2 = 0.972, RMSE = 0.035, nRMSE = 5.80%) exhibited the best accuracy in estimating LAI, followed by the SWR model (R2 = 0.765, RMSE = 0.102, nRMSE = 16.81%) and univariate linear regression model (R2 = 0.736, RMSE = 0.108, nRMSE = 17.84%). It was concluded that the estimation method based on UAV spectral parameters combined with texture features can provide an effective method for kiwifruit growth process monitoring. It is expected to provide scientific guidance and practical methods for the kiwifruit management in the field for low-cost UAV remote sensing technology to realize large area and high-quality monitoring of kiwifruit growth, thus providing a theoretical basis for kiwifruit growth investigation.

Proposing UGV and UAV Systems for 3D Mapping of Orchard Environments

During the last decades, consumer-grade RGB-D (red green blue-depth) cameras have gained popularity for several applications in agricultural environments. Interestingly, these cameras are used for spatial mapping that can serve for robot localization and navigation. Mapping the environment for targeted robotic applications in agricultural fields is a particularly challenging task, owing to the high spatial and temporal variability, the possible unfavorable light conditions, and the unpredictable nature of these environments. The aim of the present study was to investigate the use of RGB-D cameras and unmanned ground vehicle (UGV) for autonomously mapping the environment of commercial orchards as well as providing information about the tree height and canopy volume. The results from the ground-based mapping system were compared with the three-dimensional (3D) orthomosaics acquired by an unmanned aerial vehicle (UAV). Overall, both sensing methods led to similar height measurements, while the tree volume was more accurately calculated by RGB-D cameras, as the 3D point cloud captured by the ground system was far more detailed. Finally, fusion of the two datasets provided the most precise representation of the trees.

Classification of Potato Varieties Drought Stress Tolerance Using Supervised Learning

The presented study was aimed at investigating the variability for drought tolerance among potato cultivars. To achieve this, the stability of drought tolerance of potato cultivars under different water regime conditions was inspected during 11 years of consecutive experiments. The data on 50 potato cultivars’ responses to drought stress, based on the morphological features of plants, i.e., leaf and stem mass and size of the assimilation area, have been collected. The tuber yield, as well as calculated plant tolerance indexes and Climatic Water Balance for each growing season, were analyzed. The studied cultivars were later assigned into one of three tolerance groups for soil drought. The highest linear relationship was found between the mass of leaves and stems and the tuber yield but was found too weak to raise any conclusions. Thus, the ensemble learning models have been evaluated and returned better performance results, and the final classifier is the implementation of extreme gradient boosting. The final classifier of the 96.7% accuracy, which used several measured potato parameters (Relative yield decrease, Stem mass, Maturity, Assimilation area, Leaves mass, Yield per plant, calculated Climatic water balance, and indices: MSTI and DSI) that could distinguish the different tolerance groups were evaluated in the study.

Evaporation Forecasting through Interpretable Data Analysis Techniques

Climate change is increasing temperatures and causing periods of water scarcity in arid and semi-arid climates. The agricultural sector is one of the most affected by these changes, having to optimise scarce water resources. An important phenomenon within the water cycle is the evaporation from water reservoirs, which implies a considerable amount of water lost during warmer periods of the year. Indeed, evaporation rate forecasting can help farmers grow crops more sustainably by managing water resources more efficiently in the context of precision agriculture. In this work, we expose an interpretable machine learning approach, based on a multivariate decision tree, to forecast the evaporation rate on a daily basis using data from an Internet of Things (IoT) infrastructure, which is deployed on a real irrigated plot located in Murcia (southeastern Spain). The climate data collected feed the models that provide a forecast of evaporation and a summary of the parameters involved in this process. Finally, the results of the interpretable presented model are validated with the best literature models for evaporation rate prediction, i.e., Artificial Neural Networks, obtaining results very similar to those obtained for them, reaching up to 0.85R2 and 0.6MAE. Therefore, in this work, a double objective is faced: to maintain the performance obtained by the models most frequently used in the problem while maintaining the interpretability of the knowledge captured in it, which allows better understanding the problem and carrying out appropriate actions.

Evaluation of a Binary Classification Approach to Detect Herbage Scarcity Based on Behavioral Responses of Grazing Dairy Cows

In precision grazing, pasture allocation decisions are made continuously to ensure demand-based feed allowance and efficient grassland utilization. The aim of this study was to evaluate existing prediction models that determine feed scarcity based on changes in dairy cow behavior. During a practice-oriented experiment, two groups of 10 cows each grazed separate paddocks in half-days in six six-day grazing cycles. The allocated grazing areas provided 20% less feed than the total dry matter requirement of the animals for each entire grazing cycle. All cows were equipped with noseband sensors and pedometers to record their head, jaw, and leg activity. Eight behavioral variables were used to classify herbage sufficiency or scarcity using a generalized linear model and a random forest model. Both predictions were compared to two individual-animal and day-specific reference indicators for feed scarcity: reduced milk yields and rumen fill scores that undercut normal variation. The predictive performance of the models was low. The two behavioral variables “daily rumination chews” and “bite frequency” were confirmed as suitable predictors, the latter being particularly sensitive when new feed allocation is present in the grazing set-up within 24 h. Important aspects were identified to be considered if the modeling approach is to be followed up.

Classification of Daily Crop Phenology in PhenoCams Using Deep Learning and Hidden Markov Models

Near-surface cameras, such as those in the PhenoCam network, are a common source of ground truth data in modelling and remote sensing studies. Despite having locations across numerous agricultural sites, few studies have used near-surface cameras to track the unique phenology of croplands. Due to management activities, crops do not have a natural vegetation cycle which many phenological extraction methods are based on. For example, a field may experience abrupt changes due to harvesting and tillage throughout the year. A single camera can also record several different plants due to crop rotations, fallow fields, and cover crops. Current methods to estimate phenology metrics from image time series compress all image information into a relative greenness metric, which discards a large amount of contextual information. This can include the type of crop present, whether snow or water is present on the field, the crop phenology, or whether a field lacking green plants consists of bare soil, fully senesced plants, or plant residue. Here, we developed a modelling workflow to create a daily time series of crop type and phenology, while also accounting for other factors such as obstructed images and snow covered fields. We used a mainstream deep learning image classification model, VGG16. Deep learning classification models do not have a temporal component, so to account for temporal correlation among images, our workflow incorporates a hidden Markov model in the post-processing. The initial image classification model had out of sample F1 scores of 0.83–0.85, which improved to 0.86–0.91 after all post-processing steps. The resulting time series show the progression of crops from emergence to harvest, and can serve as a daily, local-scale dataset of field states and phenological stages for agricultural research.

Application of Laser-Induced Breakdown Spectroscopy Coupled With Spectral Matrix and Convolutional Neural Network for Identifying Geographical Origins of Gentiana rigescens Franch

Accurate geographical origin identification is of great significance to ensure the quality of traditional Chinese medicine (TCM). Laser-induced breakdown spectroscopy (LIBS) was applied to achieve the fast geographical origin identification of wild Gentiana rigescens Franch (G. rigescens Franch). However, LIBS spectra with too many variables could increase the training time of models and reduce the discrimination accuracy. In order to solve the problems, we proposed two methods. One was reducing the number of variables through two consecutive variable selections. The other was transforming the spectrum into spectral matrix by spectrum segmentation and recombination. Combined with convolutional neural network (CNN), both methods could improve the accuracy of discrimination. For the underground parts of G. rigescens Franch, the optimal accuracy in the prediction set for the two methods was 92.19 and 94.01%, respectively. For the aerial parts, the two corresponding accuracies were the same with the value of 94.01%. Saliency map was used to explain the rationality of discriminant analysis by CNN combined with spectral matrix. The first method could provide some support for LIBS portable instrument development. The second method could offer some reference for the discriminant analysis of LIBS spectra with too many variables by the end-to-end learning of CNN. The present results demonstrated that LIBS combined with CNN was an effective tool to quickly identify the geographical origin of G. rigescens Franch.

Over 20 Years of Machine Learning Applications on Dairy Farms: A Comprehensive Mapping Study

Machine learning applications are becoming more ubiquitous in dairy farming decision support applications in areas such as feeding, animal husbandry, healthcare, animal behavior, milking and resource management. Thus, the objective of this mapping study was to collate and assess studies published in journals and conference proceedings between 1999 and 2021, which applied machine learning algorithms to dairy farming-related problems to identify trends in the geographical origins of data, as well as the algorithms, features and evaluation metrics and methods used. This mapping study was carried out in line with PRISMA guidelines, with six pre-defined research questions (RQ) and a broad and unbiased search strategy that explored five databases. In total, 129 publications passed the pre-defined selection criteria, from which relevant data required to answer each RQ were extracted and analyzed. This study found that Europe (43% of studies) produced the largest number of publications (RQ1), while the largest number of articles were published in the Computers and Electronics in Agriculture journal (21%) (RQ2). The largest number of studies addressed problems related to the physiology and health of dairy cows (32%) (RQ3), while the most frequently employed feature data were derived from sensors (48%) (RQ4). The largest number of studies employed tree-based algorithms (54%) (RQ5), while RMSE (56%) (regression) and accuracy (77%) (classification) were the most frequently employed metrics used, and hold-out cross-validation (39%) was the most frequently employed evaluation method (RQ6). Since 2018, there has been more than a sevenfold increase in the number of studies that focused on the physiology and health of dairy cows, compared to almost a threefold increase in the overall number of publications, suggesting an increased focus on this subdomain. In addition, a fivefold increase in the number of publications that employed neural network algorithms was identified since 2018, in comparison to a threefold increase in the use of both tree-based algorithms and statistical regression algorithms, suggesting an increasing utilization of neural network-based algorithms.

The Future of Agricultural Jobs in View of Robotization

Robotics and computerization have drastically changed the agricultural production sector and thus moved it into a new automation era. Robots have historically been used for carrying out routine tasks that require physical strength, accuracy, and repeatability, whereas humans are used to engage with more value-added tasks that need reasoning and decision-making skills. On the other hand, robots are also increasingly exploited in several non-routine tasks that require cognitive skills. This technological evolution will create a fundamental and an unavoidable transformation of the agricultural occupations landscape with a high social and economic impact in terms of jobs creation and jobs destruction. To that effect, the aim of the present work is two-fold: (a) to map agricultural occupations in terms of their cognitive/manual and routine/non-routine characteristics and (b) to assess the susceptibility of each agricultural occupation to robotization. Seventeen (17) agricultural occupations were reviewed in relation to the characteristics of each individual task they entail and mapped onto a two-dimensional space representing the manual versus cognitive nature and the routine versus non-routine nature of an occupation. Subsequently, the potential for robotization was investigated, again concerning each task individually, and resulted in a weighted average potential adoption rate for each one of the agricultural occupations. It can be concluded that most of the occupations entail manual tasks that need to be performed in a standardised manner. Considering also that almost 81% of the agricultural work force is involved with these activities, it turns out that there is strong evidence for possible robotization of 70% of the agricultural domain, which, in turn, could affect 56% of the total annual budget dedicated to agricultural occupations. The presented work silhouettes the expected transformation of occupational landscape in agricultural production as an effort for a subsequent identification of social threats in terms of unemployment and job and wages polarization, among others, but also of opportunities in terms of emerged skills and training requirements for a social sustainable development of agricultural domain.

Automated Processing and Phenotype Extraction of Ovine Medical Images Using a Combined Generative Adversarial Network and Computer Vision Pipeline

The speed and accuracy of phenotype detection from medical images are some of the most important qualities needed for any informed and timely response such as early detection of cancer or detection of desirable phenotypes for animal breeding. To improve both these qualities, the world is leveraging artificial intelligence and machine learning against this challenge. Most recently, deep learning has successfully been applied to the medical field to improve detection accuracies and speed for conditions including cancer and COVID-19. In this study, we applied deep neural networks, in the form of a generative adversarial network (GAN), to perform image-to-image processing steps needed for ovine phenotype analysis from CT scans of sheep. Key phenotypes such as gigot geometry and tissue distribution were determined using a computer vision (CV) pipeline. The results of the image processing using a trained GAN are strikingly similar (a similarity index of 98%) when used on unseen test images. The combined GAN-CV pipeline was able to process and determine the phenotypes at a speed of 0.11 s per medical image compared to approximately 30 min for manual processing. We hope this pipeline represents the first step towards automated phenotype extraction for ovine genetic breeding programmes.

Smart Indoor Farms: Leveraging Technological Advancements to Power a Sustainable Agricultural Revolution

Conventional farming necessitates a large number of resources and infrastructure such as land, irrigation, manpower to manage farms, etc. Modern initiatives are required to automate conventional farms. Smart indoor farms offer the potential to remedy the shortfalls of conventional farms by providing a controlled, intelligent, and smart environment. This paper presents a three-dimensional perspective consisting of soilless farming, energy harvesting, and smart technologies, which could be considered as the three important characteristics of smart indoor farms. A six-layer smart indoor farms architecture has also been proposed, which explains how data are collected using various sensors and devices and then transmitted onto the cloud infrastructure for further analysis and control through various layers. Artificial lighting, smart nutrition management, and artificial climate control, to name a few, are some of the important requirements for smart indoor farms while considering control and service management factors. The major bottleneck in installing such systems is both the economical and the technical constraints. However, with the evolution of technology (and when they become widely available in the near future), a more favourable farming scenario may emerge. Furthermore, smart indoor farms could be viewed as a potential answer for meeting the demands of a sustainable agricultural revolution as we move closer to Agriculture 4.0. Finally, in order to adapt smart indoor farms and their study scope, our work has presented various research areas to potential researchers.

Behavioral Monitoring Tool for Pig Farmers: Ear Tag Sensors, Machine Intelligence, and Technology Adoption Roadmap

Precision swine production can benefit from autonomous, noninvasive, and affordable devices that conduct frequent checks on the well-being status of pigs. Here, we present a remote monitoring tool for the objective measurement of some behavioral indicators that may help in assessing the health and welfare status-namely, posture, gait, vocalization, and external temperature. The multiparameter electronic sensor board is characterized by laboratory measurements and by animal tests. Relevant behavioral health indicators are discussed for implementing machine learning algorithms and decision support tools to detect animal lameness, lethargy, pain, injury, and distress. The roadmap for technology adoption is also discussed, along with challenges and the path forward. The presented technology can potentially lead to efficient management of farm animals, targeted focus on sick animals, medical cost savings, and less use of antibiotics.

Bridging the Gaps in Traceability Systems for Fresh Produce Supply Chains: Overview and Development of an Integrated IoT-Based System

Traceability, namely the ability to access information about a product and its movement across all stages of the supply chain, has been emerged as a key criterion of a product’s quality and safety. Managing fresh products, such as fruits and vegetables, is a particularly complicated task, since they are perishable with short shelf lives and are vulnerable to environmental conditions. This makes traceability of fresh produce very significant. The present study provides a brief overview of the relative literature on fresh produce traceability systems. It was concluded that the commercially available traceability systems usually neither cover the entire length of the supply chain nor rely on open and transparent interoperability standards. Therefore, a user-friendly open access traceability system is proposed for the development of an integrated solution for traceability and agro-logistics of fresh products, focusing on interoperability and data sharing. Various Internet of Things technologies are incorporated and connected to the web, while an android-based platform enables the monitoring of the quality of fruits and vegetables throughout the whole agri-food supply chain, starting from the field level to the consumer and back to the field. The applicability of the system, named AgroTRACE, is further extended to waste management, which constitutes an important aspect of a circular economy.

Monitoring Soil and Ambient Parameters in the IoT Precision Agriculture Scenario: An Original Modeling Approach Dedicated to Low-Cost Soil Water Content Sensors

A low power wireless sensor network based on LoRaWAN protocol was designed with a focus on the IoT low-cost Precision Agriculture applications, such as greenhouse sensing and actuation. All subsystems used in this research are designed by using commercial components and free or open-source software libraries. The whole system was implemented to demonstrate the feasibility of a modular system built with cheap off-the-shelf components, including sensors. The experimental outputs were collected and stored in a database managed by a virtual machine running in a cloud service. The collected data can be visualized in real time by the user with a graphical interface. The reliability of the whole system was proven during a continued experiment with two natural soils, Loamy Sand and Silty Loam. Regarding soil parameters, the system performance has been compared with that of a reference sensor from Sentek. Measurements highlighted a good agreement for the temperature within the supposed accuracy of the adopted sensors and a non-constant sensitivity for the low-cost volumetric water contents (VWC) sensor. Finally, for the low-cost VWC sensor we implemented a novel procedure to optimize the parameters of the non-linear fitting equation correlating its analog voltage output with the reference VWC.

Low-Cost Wireless Sensing System for Precision Agriculture Applications in Orchards

Wireless sensor networks (WSNs) can be reliable tools in agricultural management. In this work, a low cost, low power consumption, and simple wireless sensing system dedicated for agricultural environments is presented. The system is applicable to small to medium sized fields, located anywhere with cellular network coverage, even in isolated rural areas. The novelty of the developed system lies in the fact that it uses a dummy device as Coordinator which through simple but advanced programming can receive, process, and send data packets from all End-nodes to the cloud via a 4G cellular network. Furthermore, it is energy independent, using solar energy harvesting panels, making it feasible to operate in remote, isolated fields. A star topology was followed for the sake of simplification, low energy demands and increased network reliability. The developed system was tested and evaluated in laboratory and real field environment with satisfactory operation in terms of independence, and operational reliability concerning packet losses, communication range (>250 m covering fields up to 36 ha), energy autonomy, and uninterrupted operation. The network can support up to seven nodes in a 30 min data acquisition cycle. These results confirmed the potential of this system to serve as a viable option for monitoring environmental, soil, and crop parameters.

Orchard Mapping with Deep Learning Semantic Segmentation

This study aimed to propose an approach for orchard trees segmentation using aerial images based on a deep learning convolutional neural network variant, namely the U-net network. The purpose was the automated detection and localization of the canopy of orchard trees under various conditions (i.e., different seasons, different tree ages, different levels of weed coverage). The implemented dataset was composed of images from three different walnut orchards. The achieved variability of the dataset resulted in obtaining images that fell under seven different use cases. The best-trained model achieved 91%, 90%, and 87% accuracy for training, validation, and testing, respectively. The trained model was also tested on never-before-seen orthomosaic images or orchards based on two methods (oversampling and undersampling) in order to tackle issues with out-of-the-field boundary transparent pixels from the image. Even though the training dataset did not contain orthomosaic images, it achieved performance levels that reached up to 99%, demonstrating the robustness of the proposed approach.