Scalable interactive teaching with Jupyter: environmental physics and AI solutions

Mapped principles 

The project is mapped to the following framework principles:

P5. Available experimental environments

P1. Essential knowledge of the subject  

P2. Ability to handle wicked problems for a sustainable societal development  

P3. Active student-centred learning

Definitions of the principles

Contact 

Liangchao Zou associate professor

lzo@kth.se , 087908633

Profile

Project number 

This project is implemented at the School of Architecture and the Built Environment (ABE) within the Future Education programme (project no. 2601-ABE) 

Project context 

Promoting students’ active and deep learning is crucial in higher education, especially in technical and engineering-oriented subjects such as environmental physics, where physical processes are described through complex theories and mathematical models. A higher degree of digital demonstrations (combining text, equations, code, and visualisations) and student-centred, interactive exercises could support such deeper understanding. A transition from static learning materials to interactive learning materials is necessary to ensure that students do not merely consume knowledge passively but actively work with code and simulation of the physical processes. 

In addition, even though students learn the physical laws, they currently lack the competence to integrate these with emerging data-driven methods. There is a gap between traditional analytical approaches and emerging AI methods. As environmental challenges such as climate change and resource depletion become increasingly complex (and wicked), engineers need to be able to combine these approaches. One example is how artificial intelligence (AI) can be used to handle climate dynamics and groundwater pollution. 

Since 2020, the course AE2201 Environmental Dynamics/Physical Processes has successfully used Jupyter Notebooks to illustrate physical processes. Jupyter Notebooks is a powerful interactive tool that enables teachers to create documents combining text, equations, code, and visualisations. However, the current material is spread across several separate files and lacks a structured integration of modern AI techniques.

Purpose (outcome) 

The purpose of the project is to create a coherent digital platform that is scalable and transferable to other engineering disciplines at KTH dealing with complex physical systems, by: 

• compiling existing and newly developed learning material in a single, openly accessible Jupyter Book 
• developing a new module where physical laws are integrated into neural networks, thereby combining environmental physics with AI-based solutions. 

The intended outcomes are: 

• Increased AI competence: Students at the ABE school gain highly relevant and work-related skills in using open source code and AI methods to solve difficult physical problems, positioning KTH at the forefront of modern engineering education; 
• Higher scalability: By creating an openly accessible Jupyter Book template, the project provides a reusable blueprint for other courses. This lowers the threshold for colleagues to integrate programming and AI into their subjects and promotes a culture of digital pedagogical development; 
• More sustainable learning resources: The transition to a web-based platform ensures that teaching materials become easier to update, maintain, and share compared with traditional formats. 

Project results (output) 

The project will deliver the following concrete results: 

• A new educational module on artificial intelligence for environmental dynamics (approximately 40 hours of study), with a particular focus on Physics-Informed Neural Networks (PINNs).
• An openly accessible Jupyter Book (an interactive, web-based textbook) that integrates existing course theory with the new AI/PINN applications, published in a public repository (for example GitHub). 
• A library of Python-based code templates for solving “difficult” environmental problems (for example groundwater transport and climate interactions) using both classical numerical methods and AI. 
• Guidance for KTH teachers on how the Jupyter Book structure can be adapted to other engineering courses. 
• Two workshops (one internal workshop for ABE and one open workshop for KTH) where the developed tools and pedagogical methods are presented to colleagues. 

Time plan 

Start date: 2026-03-01
End date: 2027-02-28 

Project documentation 

If you have a KTH ID, you can read the documents when logged in:

• Poster: The project’s first poster (in Swedish) will be available at the ”Storträffen” meetup autumn 2026
• Project directive (Swedish): Projektdirektiv_FrU26_2601-ABE_Jupyter-EnvAI.pdf