Postdoctoral Research Associate - Machine Learning/AI for Flow Through Porous Media

Location: South Kensington / White City

Job Summary

As a Postdoctoral Research Associate in Machine learning/AI for flow through porous media, you will be harnessing the power of AI to accelerate the energy transition and push the scientific frontier in subsurface flow through porous media.

Our generation is facing unprecedented challenges in climate change and sustainability. To prevent catastrophic climate outcomes, many countries have pledged to take rapid action to reach net zero by 2050. This target calls for a portfolio of innovative, interdisciplinary, and scalable solutions to reduce greenhouse gas emissions and accelerate the transition towards a low-carbon society. Among these solutions, subsurface geological formations play a vital and irreplaceable role in both CO2 storage and energy storage.

Through decades of research on subsurface flow, a mature methodology has been developed for characterizing complex porous media through imaging flow in porous rocks. We can use micro-CT scanners to recreate the in situ conditions of gas flowing through subsurface porous formations while "looking" inside the rock pores with micro spatial resolution. State-of-the-art research facilities can generate 3D imaging data with over 10 billion voxels several times daily, while synchrotron sources can acquire images as frequently as every second.

The Department of Earth Science Engineering at Imperial College London holds a unique advantage with access to vast amount of micro-CT rock imaging datasets, which is a gold mine for machine learning algorithms to learn from. In this position, you have the opportunity to explore and leverage this rich dataset and develop powerful state-of-the-art machine learning algorithms, simultaneously advancing subsurface flow study and AI-based approach for physics prediction. You will study the application of machine learning in image segmentation, analysis and generative modelling, including the development of physically-informed methods to simulate multiphase displacement.

Duties and responsibilities

• You will take initiative in planning research in developing innovative machine learning algorithms suitable for the porous media problems.
• You will be responsible for handling a large data repository of micro-CT imagining data of fluid flow through porous media.
• You will work with and provide guidance to PhD and MS students.
• You will collaborate with other scientists within Imperial College and elsewhere in London and abroad.
• You will submit publications and present work at geoscience/computer science conferences.
• Assist in the supervision of undergraduate and postgraduate research students and research assistants as required.
• To comply with the College, Division, and Unit safety practices and to attend courses on safety when appropriate.

Essential requirements

• Have a strong background in coding (especially cloud computing), scientific computing, and machine learning.
• You need to hold a PhD in Engineering of Physical Science, preferably in the Earth/Geo-sciences domain.
• Hold a track record of strong publications in relevant domain.
• Passionate about machine learning/AI!

Further Information

Should you require any further details on the role please contact: Gege Wen (g.wen@imperial.ac.uk).

Closing date: 6^th May 2024 

Apply online via the above ‘Apply’ button.

Advert information