진학프로석·박사 채용 접수 플랫폼

MSCA Postdoctoral Fellowship: Apply with IMT Atlantique and join us!

Thinking about your next career step and willing to boost your career? Looking to experience intersectoral, interdisciplinary and international research? Wishing to carry out your research in a place addressing industrial, environmental, societal and digital transitions as a priority challenge?

Then apply for a Marie Sklodowska-Curie (MSCA) Postdoctoral Fellowship with IMT Atlantique as your host institution!

The MSCA Postdoctoral Fellowships (PF) provide funding to promising postdoctoral researchers with a view to develop their academic and transferable skills and enhance their career prospects by moving and working abroad.

By applying with IMT Atlantique you give yourself a chance to join a technological university ranked among the top 500 universities in the world according to the Times Higher Education (THE) World University's 2026 ranking (top 301-400 for Computer Science, top 201-250 for Engineering, top 301-400 for Physical sciences, and the 124th in the 2026 THE Interdisciplinary Science ranking, and top 51-100 for Data Science and Artificial Intelligence in the 2025 Shangaï ranking. IMT Atlantique is also awarded with the “Bienvenue en France” label, a recognition delivered by Campus France for our capacity to welcome and host international researchers in the best working conditions.

Topic open:

Electromagnetic digital twins are a transformative approach to industrial design and diagnostics, enabling real-time virtual prototyping and continuous monitoring of complex systems. However, current computational electromagnetics (CEM) methods face critical bottlenecks when deployed in digital twin applications: iterative solvers require repeated solutions across parameter sweeps and design iterations, creating prohibitive computational costs; low-frequency and dense-mesh instabilities degrade accuracy in multi-scale scenarios; and ill-posed inverse problems limit reconstruction reliability under noisy or incomplete data.

This research topic addresses these challenges by developing breakthrough computational paradigms that combine three complementary innovations. First, hierarchy-free fast direct solvers based on operator filtering will enable factor-once/solve-many strategies, dramatically reducing computational overhead in multi-query scenarios typical of digital twins. Second, broadband-stable integral equation formulations using Calderón preconditioning and quasi-Helmholtz projectors will ensure robust conditioning across frequency ranges and geometric complexities. Third, physics-informed regularization techniques will stabilize electromagnetic inverse problems, enabling accurate reconstruction of material properties and defect localization from measurement data.

The candidate will focus on implementing these methods for medical and industrial diagnostics applications. The research involves deriving stabilized formulations, developing efficient compression algorithms, and validating the framework medical and industrial benchmarks. Successful outcomes will include scalable solvers achieving near-linear complexity, validated imaging pipelines for digital twin environments, and demonstration of real-time capability on representative industrial cases.

Supervisor: Associate Professor Adrien MERLINI

Keywords:Computational electromagnetics, high performance computing, inverse problems