2-Year Postdoc Position in AI–Optimization Co-Design for Materials Design
Institution: OMeGA Team, IRIMAS Institute, Université de Haute-Alsace, Mulhouse, France
Duration: 24 months, with the possibility of a 6-month extension
Funding: Part of the Green Foam PEPR project, funded by PEPR, France
Application Deadline: 30/01/2026
Project Context
Many contemporary design problems can be formulated as high-dimensional optimization tasks. The objective is to determine values for a set of decision variables so that a desired performance level is achieved under practical constraints. The primary challenge here is the prohibitive cost of evaluating a single candidate design, which may require running a complex computational workflow, performing laboratory measurements, or carrying out iterative validation steps. This expensive evaluation budget makes exhaustive search impossible and renders manual tuning impractical. Consequently, there is a critical need for intelligent optimization frameworks that leverage AI to learn from sequential evaluations, dynamically model the design space, and strategically guide the search toward high-performance regions with maximum sample efficiency.
Within the broader Green Foam project, our work package leverages this perspective by developing algorithmic methods at the interface of machine learning and optimization. The offered postdoctoral research is explicitly methodology-driven and will focus on: (1) designing AI-assisted optimization methods with learning-based adaptation mechanisms; (2) developing and exploiting surrogate/predictive models to approximate costly evaluations and guide the search; (3) using AI for dynamic selection, tuning, and adaptation of optimization strategies to improve performance across different design cases and constraints; and (4) leveraging advanced optimization techniques to train and calibrate predictive models for foam-property estimation. The proposed methods will be validated through systematic computational experiments and benchmarking against established baselines, with the goal of producing reusable algorithmic contributions and high-quality publications.
In this position, a concrete case study on polymer foam design will be used to develop, evaluate, and benchmark the proposed AI and optimization methods in a realistic setting. The position targets candidates with expertise in computer science, machine learning, optimization, or applied mathematics. No prior background in foam design is expected, since the necessary application knowledge and support will be provided through the project consortium.
Responsibilities and Tasks
Required Skills and Qualifications
About us:
OMeGA is a research team within the IRIMAS institute at Université de Haute‑Alsace in Mulhouse, France, specializing in artificial intelligence‑based optimization, metaheuristics, algorithms, and modelling for complex real‑world problems. Our work ranges from the design of new hybrid metaheuristics and AI‑driven optimization methods to their application in domains such as energy systems, electric vehicles, transportation, materials science, computer vision, and 3D geometry processing, often leveraging parallel and GPU‑based computing. More information about the team, our projects, and publications is available at https://omega-irimas.github.io/.
Application Procedure
Interested candidates should submit a cover letter detailing their research interests and relevant experience, along with a comprehensive CV, academic transcripts, Ph.D. defence reports, and two reference letters. Applications should be sent to:
· Lhassane Idoumghar: lhassane.idoumghar@uha.fr
· Mahmoud Golabi: mahmoud.golabi@uha.fr
· Abdennour Azerine: abdennour.azerine@uha.fr