GAIA: A Comprehensive Pipeline for Enabling Aircraft Digital Twin Creation

Francesco Biondani1, Luigi Capogrosso1, Uzair Khan1, Nicola Dall'Ora1, Enrico Fraccaroli1, Domenico Fabio Migliore2, Francesco Acerra2, Marco Cristani1, Franco Fummi1
1 Department of Engineering for Innovation Medicine, University of Verona, Italy 2 Leonardo S.p.A., Turin, Italy
Paper UniVR Open Access Code
🎉 Accepted @ ISIE 2025 🎉
Teaser
The four-stage GAIA pipeline for enabling the creation of aircraft digital twin. Specifically, GAIA integrates physics-driven simulations with GenAI-augmented data to improve model robustness and accuracy. Furthermore, we build DSLG D/R, a novel and accurate physics-driven dataset specifically designed for LGS fault classification in collaboration with Leonardo S.p.A.

Abstract

The integrity and reliability of Landing Gear Systems (LGSs) are crucial for aircraft safety. However, the scarcity of real-world fault data hinders the creation of effective Predictive Maintenance (PdM) strategies, especially those that rely on modern Machine Learning (ML) techniques. As a result, this article presents GAIA: the first comprehensive pipeline that enables the creation of digital twins to support PdM in the aviation domain. Specifically, combining multi-physics modeling and data-driven techniques, GAIA allows learning models to achieve superior performance in fault detection and diagnosis tasks. As a use case, we consider a LGS system, and introduce DSLG D/R, a novel dataset specifically designed for LGS fault classification, created in collaboration with Leonardo S.p.A. Our results demonstrate a significant 10.56% improvement in fault classification accuracy compared to other data augmentation methods. To further demonstrate the applicability of our method, we also evaluated it on the Electrical Fault dataset, a well-established benchmark for the diagnosis of power system faults, highlighting the GAIA versatility across different critical safety domains.

BibTeX

TBA.