Master Thesis - Optical projection models using machine learning

About Mycronic 

Mycronic is a global high-tech company whose innovative solutions have been advancing electronics technology for over 40 years, where the product lineup includes mask writers. Today we continue to grow and serve customers in an expanding variety of industries. What we do impacts the future of technology, and in turn, the way we live our lives tomorrow. 

At this moment there is no other company in the world that can compete with Mycronic's mask writers. All displays you see, whether it is your TV, laptop screen, smart watch, or infotainment screen in your car, are manufactured with the use of Mycronic's machines. The same technology is also used in the semi-conductor industry. The process for the manufacturing happens at a nanometer level, smaller than blood cells and viruses. For all of this to be possible, a great emphasis is put on the quality of software and hardware.

Background 

An important aspect in the manufacturing of lithographic masks at nanometer-scale precision is the ability to verify that the pattern-generating equipment matches the expected performance. In such verification we often use numerical simulations of physics, and we are currently looking into extending current simulation capabilities into models based on machine learning [1].

Project description 

Diploma work at master (30 hp) level, where the project will focus on the application of machine learning locked to physical principles. The project focus on a purely theoretical study of optical projection models, possibly also leaving the scalar domain and entering vector electromagnetic propagation [2]. 

The outline of the project is as follows: 

1. Study of the available literature and decision of project scope and suitable machine learning model. 
2. Implementation and training of machine learning models for optical wave propagation, following classical electrodynamics. 
3. Evaluation of model performance, with possibility of validation against finite-element models (FEM) if found feasible. 
4. Writing your MSc Thesis, followed by an oral presentation of your results. During the diploma work, there will be plenty of opportunities for getting acquainted also with experiments and ongoing product development at Mycronic.

While the work is carried out at Mycronic, the study is of a generic nature and we will strive to publish any scientific outcome in peer-reviewed academic journals. The location for the diploma work is preferably at Mycronic’s headquarters in Täby; however remote work also works perfectly well.

Your profile 

We are looking for one or two candidates who enjoy computer-assisted problem solving with state-of-the-art machine learning models. As this project is tightly associated to physical principles and Maxwell's equations, you are probably in the final stage of finishing your undergraduate studies at engineering physics (F) or electrical engineering (E) and would like to leave the path open whether to enter a corporate career or to continue with PhD studies. The scope of your project will depend somewhat on your profile and background, but it is preferable that you have a solid interest in machine learning and theory. In the project, we target implementation of the model in Python using PyTorch or TensorFlow.

Contact and further information 

Fredrik Jonsson, Core Tech and Innovation, Mycronic, fredrik.jonsson@mycronic.com; also at Ångströmlaboratoriet, Division for Electricity, Uppsala University, fredrik.jonsson@angstrom.uu.se.

References 

[1] J. Lim and D. Psaltis, MaxwellNet: Physics-driven deep neural network training based on Maxwell's Equations, Appl. Phys. Lett. Photonics 7, 011301 (2022). https://doi.org/10.1063/5.0071616 

[2] Max Born & Emil Wolf, Principles of Optics (Cambridge University Press, 2013).

A culture of collaboration and personal growth
At Mycronic, we love what we do, but most importantly who we do it with. Because to us the relationships we have with our customers and each other are the keys to success.

Take part in the excitement of working with innovative people and global businesses who are elevating today’s standards in modern electronics. Share in the responsibility of bringing great ideas to life within an inclusive culture that not only promotes personal growth and embraces diversity but depends upon it. 

Here you are expected to have a voice and will be encouraged to get involved. It’s this very mindset that empowers our people to make a positive difference for a broad range of businesses, society and the planet – every day.

Click to learn more about Mycronic and what it’s like to work with us https://www.mycronic.com/en/career/working-at-mycronic/