Biodiesel has received worldwide attention as a renewable energy resource that reduces greenhouse gas (GHG) emissions. Unlike traditional fossil fuels, such as coal, oil, and natural gas, biodiesel made of vegetable oils, animal fats, or recycled restaurant grease incurs higher production costs, so its supply chain should be managed efficiently for operational cost reduction. To this end, multiple machine learning technologies have recently been applied to estimate feedstock yield, biodiesel productivity, and biodiesel quality. This study aims to identify the machine learning technologies useful in particular areas of supply chain management by review of the scientific literature. As a result, nine machine learning algorithms, the Gaussian process model (GPM), random forest (RF), artificial neural network (ANN), support vector machine (SVM), k-nearest neighbor (KNN), AdaBoost regression, multiple linear regression (MLR), linear regression (LR). and multilayer perceptron (MLP), are used for feedstock yield estimation, biodiesel productivity prediction, and biodiesel quality prediction. Among these, RF and ANN were identified as the most appropriate algorithms, providing high prediction accuracy. This finding will help engineers and managers understand concepts of machine learning technologies so they can use appropriate technology to solve operational problems in supply chain management.

Machine vision based on artificial intelligence technology is being actively utilized to reduce defect rates in the photovoltaic module production process. This study aims to propose a machine vision approach using EfficientNet-B0 for defective photovoltaic module detection. In particular, the proposed approach is applied to the electroluminescence (EL) operation, which identifies microcracks in PV modules by using polarization current. The proposed approach extracts low-level structures and local brightness variations, such as busbars, fingers, and cell boundaries, from a single convolutional block. Furthermore, the mobile inverted bottleneck convolution (MBConv) block progressively transforms defect patterns—such as microcracks and dark spots—that appear at various shooting angles into high-level feature representations. The converted image is then processed using global average pooling (GAP), Dropout, and a final fully connected layer (Dense) to calculate the probability of a defective module. A sigmoid activation function is then used to determine whether a PV module is defective. Experiments show that the proposed Efficient-B0-based methodology can stably achieve defect detection accuracy comparable to AlexNet and GoogLeNet, despite its relatively small number of parameters and fast processing speed. Therefore, this study will contribute to increasing the efficiency of EL operation in industrial fields and improving the productivity of PV modules.

This research proposes a novel location allocation framework that utilizes agent-based simulation for the efficient production of corn stover-based bioethanol, which requires dedicated pretreatment facilities for the feedstock. The framework comprises two main modules: (1) a Pretreatment Facility Module that assesses the performance of the corn stover-based bioethanol supply chain based on the interactions among three types of agents, namely order agent, pretreatment agent, and transport agent, and (2) an Optimization Module designed to determine the optimal supply chain configuration by selecting the most suitable number and locations for pretreatment facilities to achieve the lowest total operational cost. The framework is implemented in a case study for South Korea, which aims to raise the bioethanol blending ratio from 4% in 2025 to 8% by 2030. Experimental results reveal that, within the bioethanol supply chain comprising eight farms and four refineries, a 1% increase in bioethanol blending ratio leads to an increase in the demand for approximately 2229 kL of ethanol (10,225 tons of corn stover), and the proposed framework enables to identify the optimal location of pretreatment facilities in the subject supply chain according to the change in ethanol demand.

From 2020 to 2021, crop production increased by 54% globally, and the popularity of commercial agriculture to increase profitability is gradually increasing. However, global warming and climate issues make it difficult to maintain stable crop production. To improve crop production efficiency, techniques for efficiently managing large-scale commercial farmland are needed. This study proposes a satellite image-based soil moisture and onion yield prediction technique as a methodology for managing large-scale farmland. This preemptive soil moisture management technique effectively manages increased soil pressure, resulting in soil drying due to rising temperatures. To remotely identify agricultural land, vegetation indices were extracted from satellite image data, and K-means clustering was applied. Ensemble machine learning is performed on soil images collected from satellite images. This model combines soil physical properties with soil environmental factor information to develop a model. The results show that soil color information obtained from satellite images is highly correlated with soil organic matter content. The proposed model is validated using crop yield data and environmental factor data obtained from actual crop production experiments. Consequently, the proposed methodology can be effectively applied to manage large-scale farmland and enables decision-making to improve profitability.

To achieve the goal of the 2015 Paris Agreement to limit global warming to 2 °C compared to pre-industrial levels, South Korea is implementing a policy to use bioethanol as a transportation fuel based on the Renewable Fuel Standard (RFS). This study proposes a mixed-integer linear programming (MILP) model to design an optimal bioethanol supply chain utilizing rice straw, a readily available resource in South Korea. To minimize the total cost of bioethanol production, the proposed model considers optimal facility locations, i.e., those of feedstock collection (farm), refining (refinery), and consumption (market), and transportation volumes. This experiment is conducted to evaluate the blending ratios of bioethanol in gasoline (3%, 6%, and 9%) specified by the Renewable Fuel Standard (RFS) policy, based on actual gasoline consumption data in South Korea. In the RFS 3% scenario, operating a single large-scale refinery was the most economical option, but in the RFS 6% and RFS 9% scenarios, multiple refineries must be utilized to ensure supply chain economics. In conclusion, the proposed MILP model shows the practicality of gradually increasing the number of refineries and selecting the optimal location for each region as future bioethanol demand increases.