The methodical approach of MIRELAI is based on three pillars and it integrates the physics of degradation with machine learning to enhance the reliability and predictability of electronic components and systems (ECS). Leveraging multi-scale modelling and digital twins, it aims to streamline testing, facilitate repairs, and accurately forecast electronic system lifespans.

Physics of degradation

MIRELAI Doctoral Candidates (DCs) will investigate the physics of degradation to allow for the repair of electronic components and systems (ECS) and reduce testing and verification efforts. This will be achieved by improving the current understanding of the complex physics of degradation over the entire value chain, developing accurate digital twins with a strong coupling to the product, and delivering AI-based reliability assessment tools that allow an improved and accelerated design for reliability.

Multi-scale modelling

Multi-scale modelling, and numerical simulations will be applied by the MIRELAI researchers to accurately predict the system performance. Digital twins and parameterized models will be introduced to allow for the consideration of major performance-influence factors. Simulation models at the package level, the PCB level and the system level will be elaborated. MIRELAI beneficiaries’ in-house developments and complementary commercial software will enhance the research studies.

AI-based reliability

During this stage the DCs will introduce the physics of degradation-informed machine learning, making use of the results of two other pillars to predict and improve the electronic system product reliability and predict the remaining useful lifetime (RUL). The main objective would be to train and validate surrogate models to explore massive design spaces, identify multi-dimensional correlations and manage ill-posed problems.