A design methodology to employ digital twins for remaining useful lifetime prediction in electric vehicle batteries

Predictive maintenance plays a crucial role in the context of Industry 4.0, and the adoption of Digital Twin methodologies has emerged as a promising approach for predicting the remaining useful lifetime of assets, particularly after a fault is identified. However, there is a lack of understanding regarding how to effectively apply digital twins for prognosis purposes, including estimating confidence intervals and identifying root causes of faults. To address this gap, this paper presents a methodology based on a comprehensive literature review, aiming to provide a systematic approach for predicting the remaining useful lifetime of assets. The proposed methodology encompasses several steps. It starts with data collection from physical assets or relevant databases, followed by modeling the asset’s behavior using dynamic equations. Machine learning algorithms are then applied to predict the asset’s final state in response to corrective actions. The interpretation systems provide insights and recommendations based on the virtual model, which can be further utilized by users for decision-making or implemented directly in the physical product. This methodology focuses specifically on implementing digital twins to determine the remaining lifetime of battery discharge. By following the step-by-step procedure outlined in this study, practitioners and researchers can effectively implement digital twin models and utilize them for prognosis purposes, enhancing the accuracy of predictions and enabling proactive maintenance strategies. This research contributes to bridging the gap in understanding the application of digital twins for prognosis, presenting a systematic methodology that can be applied in various domains to predict the remaining useful lifetime of assets, thereby optimizing maintenance processes and ensuring efficient asset management.