Automated Health Monitoring in Rock Drilling via Machine Learning : Industrial mining application
Zare, Marzieh (2025)
Tampere University
2025
Tieto- ja sähkötekniikan tohtoriohjelma - Doctoral Programme in Computing and Electrical Engineering
Informaatioteknologian ja viestinnän tiedekunta - Faculty of Information Technology and Communication Sciences
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Väitöspäivä
2025-10-10
Julkaisun pysyvä osoite on
https://urn.fi/URN:ISBN:978-952-03-4152-7
https://urn.fi/URN:ISBN:978-952-03-4152-7
Tiivistelmä
The mining industry significantly impacts both environmental health and worker safety. This research utilizes the potential of data fusion and machine learning methodologies to advance the monitoring and analysis of rock drilling operations.
A key aspect of our work was to develop an effective strategy for sensor placement, optimizing the collection of crucial operational data. Our primary aim was to implement and develop a robust classification system for identifying key stages of drilling operations to prevent extreme events and reduce potential damages that could lead to substantial financial losses and safety risks.
Building on our earlier research in real-time analysis techniques and operational monitoring, we designed and developed a practical machine learning model that operates as a multi-class classifier. This model uses sensory vibration signals to detect unusual patterns and irregularities, efficiently identifying potential risks. We successfully tested and validated this model by classifying five distinct mining activities, showing its effectiveness in enhancing operational efficiency and predictive maintenance capabilities.
The results of this research significantly enhance safety standards and improve operational protocols within the mining industry, demonstrating the transformative potential of integrating advanced data analysis techniques into traditional industrial practices. Our approach not only increases the accuracy of fault detection but also contributes to more sustainable mining operations by ensuring proactive risk management.
A key aspect of our work was to develop an effective strategy for sensor placement, optimizing the collection of crucial operational data. Our primary aim was to implement and develop a robust classification system for identifying key stages of drilling operations to prevent extreme events and reduce potential damages that could lead to substantial financial losses and safety risks.
Building on our earlier research in real-time analysis techniques and operational monitoring, we designed and developed a practical machine learning model that operates as a multi-class classifier. This model uses sensory vibration signals to detect unusual patterns and irregularities, efficiently identifying potential risks. We successfully tested and validated this model by classifying five distinct mining activities, showing its effectiveness in enhancing operational efficiency and predictive maintenance capabilities.
The results of this research significantly enhance safety standards and improve operational protocols within the mining industry, demonstrating the transformative potential of integrating advanced data analysis techniques into traditional industrial practices. Our approach not only increases the accuracy of fault detection but also contributes to more sustainable mining operations by ensuring proactive risk management.
Kokoelmat
- Väitöskirjat [5187]