ECO-AI Hackathon: Enabling Carbon Capture and Storage using AIPermalink

After the success of the ECO-AI Hackathon in March 2024, we’re excited to invite PhD students and PDRAs from across the UK to again join forces for three days of innovation and collaboration on AI for NetZero solutions. You’ll tackle unique challenges in carbon capture and storage and explore their economic and policy implications.

Event DetailsPermalink

Dates: Wednesday 26 March – Friday 28 March 2025
Location: Heriot-Watt University, Edinburgh, UK
Venue: Scott Suite, Edinburgh Business School

ApplicationPermalink

This in-person event will include informative workshops and mentorship led by experts from Heriot-Watt University and Imperial College London involved in the ECO-AI project. Successful applicants will receive £300 in funding toward travel and accommodation expenses.

Application Deadline: Monday 27 January 2025 (now closed)

What to ExpectPermalink

This is a two-parts event, including the ECO-AI project workshop (24-25 March), where you’ll attend presentations and panel discussions on using AI for CO₂ capture and storage. Following this, the Hackathon (26-28 March) will involve team-based work on one of three challenge tracks:

• CO₂ Storage (poreflow challenge): Fast prediction of two phase flow in porous media using deep learning
• CO₂ Capture (amineprop challenge): Predicting properties of amine molecules for carbon capture using AI
• CCS policy and economics (policyalign challenge): LLM-Driven analysis of climate policies and SDG alignment

Started codes and challenges outlinePermalink

• github repository for all challenges (URL, large files removed)
• poreflow challenge on Kaggle (starter notebook; dataset)
• amineprop challenge on Kaggle (starter notebook; dataset)
• policyalign challenge on Kaggle (starter notebook; description &dataset)

Event FlowPermalink

1. Attend 2-day AI workshop on CCS
2. Form teams and select challenge track
3. Collaborate on solutions over 3-day hackathon
4. Demo solutions to experts from industry and academia

Participants will have access to all necessary platforms and software for deep learning experiments throughout the Hackathon. This is a unique opportunity to collaborate across institutions, network with participants, and showcase your work to data scientists and industry experts.

Food and beverages will be provided throughout the event

Background & pre-readingPermalink

All participants are expected to possess basic knowledge of Python and some ML tools (e.g., Scikit-learn, PyTorch). For those who are not familiar with these tools, pre-reading on how to use basic Python, NumPy, Pandas, Scikit-learn, and PyTorch using the following resources is recommended:

1. Enthought Numpy Tutorial - SciPyConf 2023
2. University of Washington Machine Learning with Python Course
3. PyTorch Tutorials

Tentative Hackathon Agenda:

Contact UsPermalink

If you have any questions about this event, please contact ECO-AI project manager at aw2073@hw.ac.uk

Hackathon photos (March 2025)Permalink