Kasper Jul Lassen
University
Aalborg University (AAU)
Company
Novonesis A/S
Project title
Modelling framework for an Ultrafiltration Unit using Physics Informed Neural Network
Project description
This project develops a Physics-Informed Neural Network (PINN) model for an ultrafiltration unit, aiming to optimize membrane flux and reduce energy in enzyme recovery. Unlike traditional dynamic models, which struggle with phenomena like fouling, and standard neural networks, which lack physical consistency, PINNs integrate domain knowledge and machine learning. They enforce physical laws, ensuring data-driven yet physically accurate predictions. Advantages include robust predictions of fouling, adaptability, and physical interpretability, enhancing decision-making and operational efficiency.
What was your motivation for becoming a Helix Lab Fellow?
I see it as a great opportunity for me to make contacts in the industry and a chance to meet other fellow students and make new friends. Many of the fellowships programs also sounds interesting and I would like to participate in them. Lastly, moving to Kalundborg will allow me to see a new part of Denmark that I have never experienced before.
What is the most exciting part of being a Helix Lab Fellow?
One area I am very excited about is to work in close collaboration with Novonesis and to get a real feeling for how it is to work at a real company. I also really excite abut all the resources the helix lab fellows provides like access to labs and weekly events that helps develop knowledge about the city, and the people and companies with in it.
What perspectives do you see for your future after being a Helix Lab Fellow?
I'm interested in exploring the potential of the Kalundborg industry following my graduation. The innovative and sustainable practices implemented through programs like the Kalundborg Symbiosis align closely with my professional interests and career goals and are something I can see myself being a part of.