Project: Pattern discovery to improve overlay control loop by using Bayesian inference tooling
Description
This assignment aims to detect and quantify persistent overlay improvements by investigating a larger data set systematically.
It will provide you with insights into the overlay performance of ASML lithography machines. You will also learn how ASML maintains machine performance via drift control strategy. As for hands-on experience, you will deploy your MATLAB skills to analyze mathematical models and statistical toolbox, to interpret real machine population data and to link them to a physical mechanism by talking to experts.
Your tasks will include writing a test plan, collecting and preprocessing appropriate data sets, applying advanced machine learning and statistical tools to the processed data sets with various model settings, performing proof book analysis, and presenting and writing reports.
The goal is to help the overlay integration group in choosing the path to follow with respect to potential future overlay control product improvements: do we expect benefit in changing the current overlay control model; if so, which part of the current model can be improved, and what will the impact be?
Your profile
- Strong math or physics background
- Strong machine learning and data science skills
- Affinity with all or some of the following: statistical analysis, Bayesian inference, Gaussian Processes
- Knows how to program in Matlab and use Git
- (Aspiring) Cum laude student
This is a graduation internship for 5 days a week with duration of a minimum 6 months. You will need to interview with ASML before being able to start the project. (See attached information.)
Detailed description
View PDFDetails
- Supervisor
-
Erik Quaeghebeur
- Secondary supervisor
-
SRSejong Park, Hamideh Rostami
- External location
- ASML
- Interested?
- Get in contact