PhD: Robust Optimisation of Complex Optical Systems for Imaging
Where: Leiden University, 
Netherlands
What:
Within the NWO AI4optics project, this PhD develops robust and explainable optimisation methods for advanced imaging optics. AI4optics brings together experts in computer science, mathematics and physics to develop AI-based methods for solving inverse problems in high-tech optical systems. The project aims to make optical simulation, design and reconstruction easier using AI, ultimately contributing to smaller, more powerful and energy-efficient integrated circuits, reduced light pollution and improved Earth observation from space.
Traditional optical design workflows often produce systems that perform well under ideal conditions but degrade significantly due to manufacturing imperfections, alignment errors or environmental variations. In this project, you will develop robust multi-objective optimisation algorithms that explicitly account for these uncertainties during the design process. By combining evolutionary and hybrid optimisation with uncertainty quantification, you will identify not just optimal but stable optical designs. The research includes mixed-variable optimisation problems involving both continuous and discrete design parameters, as well as competing objectives such as image quality versus cost and compactness versus stray light.
A second research direction focuses on explainable AI for optical design. You will develop methods that make optimisation more transparent by revealing which design parameters most strongly influence performance and robustness. Building on recent advances in explainable AI, surrogate modelling and sensitivity analysis, you will investigate how to explain optimisation outcomes, identify the causes of design failures and provide actionable insights to optical designers and engineers.
Who: Anna Kononova, a.kononova@liacs.leidenuniv.nl
When: Until 2026-09-01 05:59
Presented at next GECCO?: yes