An RUL-informed approach for life extension of high-value assets

Abstract

The conventional approaches for life-extension (LE) of industrial assets are largely qualitative and focus only on a few indicators at the end of an asset’s design life. However, an asset may consist of numerous individual components with different useful lives and therefore applying a single LE strategy to every component will not result in an efficient outcome. In recent years, many advanced analytics techniques have been proposed to estimate the remaining useful life (RUL) of the assets equipped with sensor technology. This paper proposes a data-driven model for LE decision-making based on RUL values predicted on a real-time basis during the asset’s operational life. Our proposed LE model is conceptually targeted at the component, unit, or subsystem level; however, an asset-level decision is made by aggregating information across all components. Consequently, LE is viewed and assessed as a series of ongoing activities, albeit carefully orchestrated in a manner similar to operation and maintenance (O&M). The application of the model is demonstrated using the publicly available NASA C-MAPSS dataset for large commercial turbofan engines. This approach will be very beneficial to asset owners and maintenance engineers as it seamlessly weaves LE strategies into O&M activities, thus optimizing resources.