A Novel High-Level Reasoning Architecture for Aircraft Prescriptive Maintenance
TL;DR
An innovative High-Level Reasoner (HLR) decision support architecture for aircraft systems that provides a reusable framework for extension across aircraft systems and wider IVHM applications and serves as an enabling technology that advances beyond existing diagnostics and prognostics solutions for asset utilisation.
Abstract
State-of-the-art Integrated Vehicle Health Management (IVHM) systems and digital twins (DTs) integrate physics-based and data-driven methodologies for predictive maintenance. Such systems commonly incorporate multiple DT instances to estimate substantive outputs. Nonetheless, they exhibit key limitations: lack of multi-DT orchestration mechanisms, limited uncertainty quantification, and insufficient prescriptive decision-support capability. To this end, this paper introduces an innovative High-Level Reasoner (HLR) decision support architecture for aircraft systems. The proposed HLR architecture comprises a multi-layer data-transfer structure, with the principal HLR layer consisting of adaptable specialist modules that facilitate a robust decision support implementation for query-driven prescriptive maintenance. The developed architecture is illustrated on an aircraft landing gear system (ATA 32), orchestrating multiple federated subsystems; represented by the Brake Temperature DT and Tyre Pressure DT. The contribution is a modular architecture that provides a reusable framework for extension across aircraft systems and wider IVHM applications. It therefore serves as an enabling technology that advances beyond existing diagnostics and prognostics solutions for asset utilisation.