المعهد التقني التكنولوجي

First Research Paper

Condition Monitoring Model Development for Belt Systems Using Hybrid CNN–BiLSTM Deep-Learning Techniques

This study presents the development of a hybrid deep learning model that integrates Convolutional Neural Networks (CNN) with Bidirectional Long Short-Term Memory (BiLSTM) networks to analyze belt drive systems and predict their operational condition. The proposed model demonstrated exceptional predictive performance, achieving an accuracy of approximately 99% in detecting faults and imbalance conditions. This represents a significant advancement toward the implementation of intelligent predictive maintenance systems, contributing to the reduction of unexpected failures and production losses.

Second Research Paper

Surface Roughness Prediction in Turning Stainless Steel Applying Deep Learning and LSTM Networks

This research was presented at the:
7th International Conference on Industry of the Future and Smart Manufacturing – Malta

The study focuses on the application of Long Short-Term Memory (LSTM) networks to predict surface quality in stainless steel turning processes. The model relies on the analysis of vibration signals alongside operational parameters. The results demonstrated a high level of prediction accuracy, highlighting the model’s effectiveness in improving product quality across industrial applications.

Notably, the deep learning models were trained using high-performance GPU computing servers at Otto von Guericke University Magdeburg, Germany, enabling accelerated training processes and efficient data analysis. This significantly enhanced the accuracy and reliability of the obtained results.

This achievement represents a valuable scientific contribution that supports the integration of Artificial Intelligence technologies in the development of industrial monitoring systems and the optimization of production efficiency.