One-third of Thailand’s workers are in agriculture, but the country’s agricultural GDP is still less than 10% of its total GDP. Most Thai farmers are smallholders with limited land and low incomes. To improve the agricultural GDP and the economic situation of smallholder farmers, the Thai Government has been trying for decades to encourage and support smallholder farmers to adopt modern farming methods and smart farming equipment, including digital technologies. However, the improvement is still sluggish due to a lack of an effective approach to delivering essential digital knowledge and skills, as well as investment support for smart farming equipment. These have hindered smallholder farmers’ digital farming skill progress. To address this issue, the Broadcasting and Telecommunications Research and Development Fund for Public Interest has funded a project to develop the Digital Farmer Development Framework. This framework provides essential digital knowledge, training, coaching, and fundamental resources to upgrade smallholder digital-farming literacy to become digital farmers using problem- or project-based learning approaches and collaborative blended learning theories. Bloom’s taxonomy is used as a guideline for evaluating the framework’s effectiveness. Implementation of the Digital Farmer Development Framework has shown that farmers can significantly improve their digital farming literacy and are capable of using digital technology to improve farm management and productivity. Based on Bloom classification guidelines, 100% of the farms in the project can apply digital skills and utilize fundamental smart farming equipment as well as able to evaluate and analyze data from IoT devices. Moreover, 66% can create their own smart-system solution from fundamental smart farming tools for their farm. The project has also created a digital farmer community that shares knowledge and resources with others.

Temperature and humidity excursions during transport accelerate quality loss in fresh produce. This study develops and validates a self-contained Internet of Things (IoT) platform for in-transit monitoring and energy-aware operation. The battery-powered device operates independently of vehicle power and continuously logs temperature, relative humidity, GPS position, and onboard power draw. Power budgeting combines firmware-level deep-sleep scheduling with a LiFePO₄ battery pack, enabling uninterrupted operation for up to four days. Using ∼10,000 time-stamped observations collected over four consecutive days in a standard dry truck (non-commercial validation), we trained and compared Gradient Boosting Machine (GBM), Random Forest (RF), and Linear Regression (LR) models to predict energy consumption under varying environmental and routing conditions. GBM and LR achieved high explanatory power ($R^2\approx 0.88$) with a mean absolute error of 0.77 A·h, while RF provided interpretable feature importance data, identifying temperature as the dominant driver, followed by trip duration and humidity. The end-to-end system integrates an EMQX MQTT broker, a Laravel web application, MongoDB storage, and Node-RED flows for real-time dashboards and multi-day historical analytics. The proposed platform supports proactive decision-making in perishable logistics, with the AI analysis validating that the collected time-aligned on-route data can configure sampling/transmit cadence to preserve autonomy under stressful conditions.

Educational data mining is the method for extracting and discovering new knowledge from education data. As education data is often complex and imbalanced, it requires a data preprocessing step or learning algorithms in order to obtain accurate analysis and interpretation. Many studies emphasize on classification and clustering methods in order to get insight and comprehensive knowledge from education data. However, a small number of previous works exclusively focused on the preprocessing of education data, particularly on the topic of the imbalanced dataset. Therefore, this research objective is to enhance the accuracy of data classification in educational web usage data. Our study involves the application of synthetic minority over-sampling techniques (SMOTE) to preprocess the raw dataset from web usage data. The minority class is a group of the students who failed the examination and the majority class is the students who passed the examination. In our experiments, four synthetic minority over-sampling methods are applied, SMOTE, and its variants: Borderline-SMOTE1, Borderline-SMOTE2, and SVM-SMOTE, in order to balance the number of samples in the minority class. The experiments are evaluated by comparing the results from well-known classification methods that are Naive Bayesian, decision tree, and k-nearest neighbors. The study experiments with real-world datasets from education data. The results present that synthetic minority over-sampling methods are capable of improving the detection of the minority class and achieve improving classification performance on precision, recall, and F1-value. Ed