European Training Network on Grey-Box Models for Safe and Reliable Intelligent Mobility Systems

Projektinformationen

GREYDIENT

ID Finanzhilfevereinbarung: 955393

DOI

10.3030/955393

Projekt abgeschlossen

EK-Unterschriftsdatum 19 August 2020

Startdatum 1 Januar 2021

Enddatum 31 Dezember 2024

Finanziert unter

• EXCELLENT SCIENCE - Marie Skłodowska-Curie Actions

Gesamtkosten

€ 3 938 269,68

EU-Beitrag

€ 3 938 269,68

Koordiniert durch

KATHOLIEKE UNIVERSITEIT LEUVEN
Belgium

Periodic Reporting for period 2 - GREYDIENT (European Training Network on Grey-Box Models for Safe and Reliable Intelligent Mobility Systems)

Berichtszeitraum: 2023-01-01 bis 2024-12-31

Zusammenfassung vom Kontext und den Gesamtzielen des Projekts

Every engineered system is designed based on expected conditions and lifespan, yet uncertainties arise due to deviations in real-world operation. These uncertainties can lead to unexpected failures, impacting safety, efficiency, and long-term sustainability.

Traditional approaches to managing uncertainty rely on excessive safety margins, which, while increasing reliability, also drive up costs and resource consumption. A more refined approach is uncertainty quantification, which evaluates the probability of failure by modeling variations in design and operational conditions. However, this method encounters two main challenges: the high computational cost of running complex simulations and the limited accuracy of models relying on either purely physical principles or historical data.

To address these limitations, the GREYDIENT project develops grey-box modeling, a hybrid approach that combines physics-based (white-box) and data-driven (black-box) models. By integrating the strengths of both, grey-box models enhance predictive accuracy while reducing computational demands. This approach has broad applications in energy grids, transportation, battery management, and manufacturing, enabling more precise reliability assessments, cost-effective design optimizations, and improved system monitoring.

Through the collaboration of 10 academic and industrial partners, GREYDIENT supports 15 PhD researchers in pushing the frontiers of uncertainty quantification, predictive modeling, and robust optimization strategies for complex engineering systems.

Arbeit, die ab Beginn des Projekts bis zum Ende des durch den Bericht erfassten Berichtszeitraums geleistet wurde, und die wichtigsten bis dahin erzielten Ergebnisse

Project Management & Training (WP4, WP6, WP7):
GREYDIENT was efficiently managed through structured governance, including biannual supervisory board meetings and biweekly ESR check-ins. Recruitment challenges were addressed, ensuring minimal disruptions. A comprehensive training program was delivered, equipping researchers with skills for both academia and industry.

Scientific Advancements (WP1-WP3):

+ WP1: Developed novel multi-fidelity modeling frameworks, integrating experimental data and simulations for accurate uncertainty quantification.

+ WP2: Applied grey-box methods to structural reliability, vehicle safety, and equipment durability, improving optimization under uncertainty.

+ WP3: Implemented grey-box Digital Twins for real-time monitoring and predictive maintenance in energy grids, electric vehicle batteries, and nuclear systems.

Dissemination & Impact (WP5):

+ 20+ conference presentations at ECCOMAS, ESREL, ISMA-USD, etc.

+ 17+ journal publications in top-tier journals.

+ Public engagement via social media, including 30 pitch videos on YouTube, LinkedIn outreach (650+ followers), and a Zenodo research repository.

The GREYDIENT consortium continues to strengthen its research dissemination and industry collaboration, ensuring lasting impact beyond the project’s completion.

Fortschritte, die über den aktuellen Stand der Technik hinausgehen und voraussichtliche potenzielle Auswirkungen (einschließlich der bis dato erzielten sozioökonomischen Auswirkungen und weiter gefassten gesellschaftlichen Auswirkungen des Projekts)

Scientific & Technological Contributions:

+ Multi-fidelity modeling techniques optimizing trade-offs between computational efficiency and predictive accuracy.

+ Enhanced risk monitoring, applied in nuclear safety, vehicle crashworthiness, and manufacturing reliability.

+ Grey-box Digital Twins enabling real-time system monitoring and adaptive decision-making.

Future Directions:

+ Expanding uncertainty quantification frameworks, with a collaborative scientific publication planned.

+ Strengthening industry-academia partnerships to ensure practical applications.

+ Enhancing open-access data sharing, supporting long-term research accessibility.

Socio-Economic & Environmental Impact:

+ Cost reduction & industrial efficiency: Bridging research with real-world applications to optimize design and reliability.

+ Sustainability: Applications in renewable energy and battery health monitoring improve efficiency and reduce waste.

+ Public awareness: Outreach efforts help communicate the importance of engineering reliability and uncertainty management to a broader audience.

Conclusion: GREYDIENT’s pioneering grey-box methodologies are shaping the future of intelligent, reliable, and sustainable engineering, with far-reaching benefits for industry and society.

Simplified overview of the grey-box approach