Digital Twin Development for Feed Drive Systems Condition Monitoring and Maintenance Planning

Abstract

Current Prognosis and health management (PHM) technology suffers from challenges such as data availability, system interoperability, scalability, and transferability. In previous years, the PHM field has advanced a lot, but very few studies have been presented in which these challenges are addressed, and hence, PHM solutions are still confined to the lab environment. Digital Twin technology has the potential to address these challenges altogether and can add significant value to the PHM field. This thesis aims to develop an implementable Digital Twin framework for feed drive systems' condition monitoring and maintenance optimization, targeting these prevalent PHM challenges. The proposed framework will employ multiple physics-based models to generate synthetic data for different system states, configurations, and applications, and utilize this data with the help of machine learning to overcome the PHM challenges. The successful address of these challenges will pave the foundation in the direction of generalization of PHM solutions and also enhance the trustworthiness and reliability of PHM solutions.