Publication Search

Background: The development of industrial technology in the Industry 4.0 era has encouraged the implementation of intelligent monitoring systems to improve machine reliability and operational efficiency. However, machine fault diagnosis systems based on artificial intelligence often face limitations in terms of interpretability because the models used are complex and difficult to explain. Objective: This study aims to develop a deep learning-based industrial machine fault diagnosis system integrated with an Explainable Artificial Intelligence (XAI) approach to improve diagnostic accuracy while providing interpretable insights for users. Method: The research method involves collecting data from industrial machine sensors consisting of vibration signals, temperature measurements, and acoustic signals, followed by data preprocessing and feature extraction processes. The processed data are then used to train a deep learning-based diagnostic model, after which explainability methods such as SHAP or LIME are applied to analyze the contribution of each feature to the model’s prediction results. Model performance is evaluated using accuracy, precision, recall, and F1-score metrics. Results: The results indicate that the proposed deep learning model achieves better performance compared to conventional machine learning methods such as Support Vector Machine and Random Forest. Furthermore, the explainability analysis reveals that vibration amplitude, increases in machine component temperature, and anomalies in acoustic signals are the main factors influencing machine fault detection. Therefore, the proposed system not only improves the accuracy of machine fault diagnosis but also provides transparency in the decision-making process, thereby supporting the implementation of predictive maintenance in smart manufacturing environments.

This study investigates the long-term impact of adaptive digital learning ecosystems on students' self-regulated learning (SRL) behaviors and academic persistence. Adaptive learning systems personalize the learning experience by adjusting content and feedback to meet individual students' needs, preferences, and performance. These systems enhance engagement, motivation, and learning outcomes through real-time adjustments and continuous feedback. The research aims to explore how adaptive learning systems influence SRL and academic persistence in university courses over time. Using a longitudinal quantitative design, the study tracks SRL behaviors and academic persistence at multiple points during the semester. Results show significant improvements in SRL behaviors such as goal setting, planning, self-monitoring, and reflection among students engaged with adaptive learning environments. These students exhibited greater autonomy, improved metacognitive awareness, and higher motivation. Additionally, students in adaptive systems demonstrated greater academic persistence, as indicated by more time spent on tasks, higher assignment completion rates, and sustained engagement. The findings suggest that adaptive learning platforms promote SRL and academic persistence by offering personalized, responsive learning experiences. Unlike static, non-adaptive environments, adaptive systems provide dynamic support, enhancing students' ability to regulate their learning and remain engaged despite challenges. The study concludes that adaptive learning systems are vital for long-term academic success, though further research is needed to assess the sustainability of these effects in various educational settings and among diverse student populations.

Background: The rapid advancement of digital technologies in the Industry 4.0 era has transformed industrial mechanical systems into highly interconnected and data driven environments through the integration of sensors, the Internet of Things (IoT), data analytics, and cyber physical systems. This increasing complexity requires more adaptive and accurate monitoring and prediction methods than conventional simulation approaches, which often face limitations in capturing real time dynamic system behavior. Objective: This study aims to develop a predictive performance model for industrial mechanical systems by integrating Digital Twin technology with Physics Informed Machine Learning in order to improve monitoring accuracy and support predictive maintenance strategies. Methods: This research adopts a data driven modeling and simulation approach by developing a digital representation of an industrial mechanical system that is connected to real time sensor data. The prediction model is constructed using a Physics Informed Neural Network (PINN), which integrates operational data with physical principles governing system dynamics. The research process includes the development of a Digital Twin model, integration of sensor data, training of the PINN model, model validation using experimental data, and evaluation of prediction performance using statistical metrics. Results: The results indicate that the integration of Digital Twin technology and PINN significantly improves the prediction accuracy of industrial mechanical system performance compared with conventional simulation methods and purely data driven machine learning models. The proposed model is capable of representing system dynamics more consistently, accurately following sensor data patterns, and providing strong potential for supporting machine condition monitoring and predictive maintenance strategies in modern industrial environments.

Background: Structural Health Monitoring plays a critical role in ensuring the safety, reliability, and sustainability of high performance composite structures used in aerospace, civil infrastructure, and mechanical systems. Conventional externally mounted sensors often face challenges related to environmental interference, maintenance complexity, and long term stability. Objective: This study aims to develop and validate an integrated smart composite monitoring system with embedded sensing capabilities that enhances damage detection accuracy and operational durability under varying mechanical stress conditions. Method: Smart composite specimens were fabricated by embedding fiber optic and piezoelectric sensors within fiber reinforced polymer laminates, followed by tensile, fatigue, and vibration testing. Signal processing techniques including time frequency analysis were applied to extract damage sensitive features, which were then classified using machine learning algorithms to distinguish healthy and damaged structural states. Results: The experimental findings demonstrate high damage detection capability, stable sensor performance under cyclic loading, improved reliability compared to conventional monitoring approaches, and consistent monitoring accuracy throughout the fatigue life of the specimens. The integration of embedded sensing and data driven analytics significantly enhances structural response interpretation and supports predictive maintenance strategies.

Background: The rapid advancement of Industry 4.0 has accelerated the integration of digital technologies such as the Industrial Internet of Things (IIoT), edge computing, and Digital Twin systems in smart manufacturing environments. However, many existing implementations remain fragmented and heavily dependent on centralized cloud infrastructures, resulting in latency constraints, limited scalability, and suboptimal real-time decision making. Objective: This study aims to develop and validate an integrated edge based Digital Twin optimization framework that combines IIoT sensing, hybrid edge cloud architecture, and reinforcement learning based adaptive control. Methods: The research adopts a multi phase design consisting of framework development, simulation based validation, and industrial pilot implementation. The proposed system integrates real time data acquisition, localized edge processing, Digital Twin synchronization, and intelligent optimization mechanisms to enhance operational efficiency. Results: The findings demonstrate significant performance improvements compared to conventional cloud based systems, including substantial latency reduction, increased production throughput, reduced downtime, and improved energy efficiency. Scalability and robustness testing further confirm that distributed edge intelligence enhances system resilience under increased workloads and network disruptions. These results indicate that integrating edge computing with Digital Twin modeling and reinforcement learning provides a scalable, responsive, and energy efficient solution for next-generation smart factories.

Background: Additive manufacturing (AM) requires reliable and efficient defect detection mechanisms to ensure structural integrity and product quality, yet conventional inspection approaches remain time-consuming and often unsuitable for real-time industrial deployment. Objective: This study aims to develop and experimentally validate an artificial intelligence based vision inspection system capable of accurately detecting surface defects in AM components. Methods: A Convolutional Neural Network (CNN) architecture utilizing pretrained backbones (ResNet and EfficientNet) was implemented with a transfer learning strategy and data augmentation techniques. High-resolution AM surface images representing porosity, cracks, and layer misalignment were used for training and evaluation. Model performance was assessed using Accuracy, Precision, Recall, F1-score, and mean Average Precision (mAP), and comparative benchmarking was conducted against traditional machine learning models such as Support Vector Machine and Random Forest. Results: The proposed CNN-based models significantly outperformed conventional approaches, achieving up to 95.1% Accuracy and 92.8% mAP. The EfficientNet backbone demonstrated superior generalization capability, particularly in balancing Precision and Recall, indicating robust defect detection performance across multiple categories. These findings confirm that AI-driven inspection frameworks provide scalable and reliable quality assurance solutions for advanced manufacturing environments.