Arjun Nepal

In my thesis, I developed and implemented a Machine Learning-based system for real-time tool wear detection, monitoring, and Remaining Useful Life (RLU) prediction in CNC machining. The project involved creating a software solution that integrated sensor data to enhance production efficiency, prevent failures, and reduce costs. I independently handled all stages, including system development, algorithm implementation, performance validation, and comprehensive documentation.

I developed and deployed applications using .NET Web API and Blazor, integrated internal Inventory Management with SAP, and enhanced production through automated quality control using computer vision. Additionally, I implemented large-scale data integration from SCADA systems to Azure Data Factory using ETL processes.

I optimized operational efficiency by replacing resource-intensive Discrete Event Simulation (DES) with deep learning models, performed statistical analysis to evaluate the impact of training sample size on model performance, and developed an algorithm to determine the optimal sample size for short-term manufacturing simulations. Additionally, I collaborated with cross-functional teams to enhance strategic decision-making and operational control.