PHD
· AMSTERDAM, Netherlands ·
- Employment
- Contract
- Short-Wavelength Light Sources for EUV MetrologyAngana Mondal - Materials department - Contact DynamicsBart Weber - Materials & Surface Science for EUVLRoland Bliem - Materials Theory and ModelingEmilia Olsson PhD: Physics-Informed Machine Learning for Semiconductor Metrology Work Activities
How can we combine machine learning and physics to recover nanoscale information from imperfect images?
Modern computer chips are built with features only a few nanometers across, yet manufacturers need to measure these structures with extraordinary precision and do so quickly enough to keep up with large-scale production. This creates a fascinating computational challenge: how can we infer hidden physical properties from limited, noisy, and low-resolution measurement data?
In this project, you will develop a novel physics-informed machine learning approach that integrates physical simulations of the measurement process with its inverse reconstruction. A key challenge is the data-driven design of the experimental setup: exploring how the choice of measurements and configurations can be optimized to extract the most useful information for reliable parameter reconstruction.
You will work in close collaboration with the research department at ASML, the Centrum Wiskunde & Informatica (CWI, Prof. dr. Tristan van Leeuwen), and the AI4Science Lab, Informatics Institute, University of Amsterdam (dr. Patrick Forré), combining industrial relevance with academic depth in computational science and mathematical modeling.
Qualifications
You have (or soon will have) a MSc degree in computer science, machine learning, artificial intelligence, applied mathematics, physics, or a related discipline, meeting the Dutch university requirements for entry into a PhD program. A background in machine learning, inverse problems, scientific computing, or related data-driven methods is highly desirable. You are curious about combining physical modeling with data-dri...
How can we combine machine learning and physics to recover nanoscale information from imperfect images?
Modern computer chips are built with features only a few nanometers across, yet manufacturers need to measure these structures with extraordinary precision and do so quickly enough to keep up with large-scale production. This creates a fascinating computational challenge: how can we infer hidden physical properties from limited, noisy, and low-resolution measurement data?
In this project, you will develop a novel physics-informed machine learning approach that integrates physical simulations of the measurement process with its inverse reconstruction. A key challenge is the data-driven design of the experimental setup: exploring how the choice of measurements and configurations can be optimized to extract the most useful information for reliable parameter reconstruction.
You will work in close collaboration with the research department at ASML, the Centrum Wiskunde & Informatica (CWI, Prof. dr. Tristan van Leeuwen), and the AI4Science Lab, Informatics Institute, University of Amsterdam (dr. Patrick Forré), combining industrial relevance with academic depth in computational science and mathematical modeling.
Qualifications
You have (or soon will have) a MSc degree in computer science, machine learning, artificial intelligence, applied mathematics, physics, or a related discipline, meeting the Dutch university requirements for entry into a PhD program. A background in machine learning, inverse problems, scientific computing, or related data-driven methods is highly desirable. You are curious about combining physical modeling with data-dri...