DrSam Cooper
Associate Professor in Machine Learning for Materials Design
Dyson School of Design Engineering - Faculty of Engineering
Orcid identifier0000-0003-4055-6903 (opens in a new tab)
- Associate Professor in Machine Learning for Materials DesignDyson School of Design Engineering - Faculty of Engineering
- Dyson Building, South Kensington Campus, United Kingdom
BIO
Sam is an Associate Professor in Artificial Intelligence for Materials Design in the Dyson School of Design Engineering and leader of the "Tools for Learning, Design, and Research" (TLDR) group. His work is primarily focused on the application of AI to the design of materials for next generation energy storage technologies. In 2023 he spun-out a company, Polaron, along with two of his former PhD students, Isaac Squires and Steve Kench (www.polaron.ai), where he is the Chief Scientist. In 2025, Polaron was awarded the £1M inaugeral Manchester Prize in "AI for the Public Good".
He is currently the lecturer for the first year Engineering Mathematics course, as well as the departmental Admissions Tutor. Sam has built an online course on Multivariate Calculus for the Coursera specialisation Mathematics for Machine Learning. The specialisation has had >600,000 enrollments since it launched in 2018.
Sam originally joined Imperial in 2008 to study an undergradute degree in Mechanical Engineering, after which he undertook a PhD in the Department of Materials under the supervision of Professor John Kilner on the characterisation of solid oxide fuel cell materials. Following this, he was a postgraduate research associate in the Electrochemical Science and Engineering group led by Professor Nigel Brandon, involved with the Energy Storage for Low Carbon Grids project. In 2017 he started his permanent position in the Dyson School of Design Engineering where he founded the TLDR group.
Sam's group have recently developed a method for using Generative AI to optimise the microstructure of advanced materials such as battery electrodes, alloys, and cements. The group have also published a method for generating 3D samples from 2D image data, which is vital for a variety of materials characterisation activities.
Sam and his team have developed several open-source packages that have gathered a diverse user base within the materials microstructure community, which can all be accessed through the group's github page.
A few well-cited examples that are applicable to a wide range of materials:
* TauFactor is a tool for the rapid analysis of 3D microstructural data derived for various tomography techniques: https://www.theoj.org/joss-papers/joss.05358/10.21105.joss.05358.pdf
* SliceGAN is a AI-based model for generating 3D microstructural data from a single 2D image: https://www.nature.com/articles/s42256-021-00322-1
* ImageRep is a method for calculating the representativity of a microstructural metric from a single image: www.imagerep.io
* SambaSegment is an image segmentation tool that makes use of Meta's Segment Anything Model: www.sambasegment.com
He is currently the lecturer for the first year Engineering Mathematics course, as well as the departmental Admissions Tutor. Sam has built an online course on Multivariate Calculus for the Coursera specialisation Mathematics for Machine Learning. The specialisation has had >600,000 enrollments since it launched in 2018.
Sam originally joined Imperial in 2008 to study an undergradute degree in Mechanical Engineering, after which he undertook a PhD in the Department of Materials under the supervision of Professor John Kilner on the characterisation of solid oxide fuel cell materials. Following this, he was a postgraduate research associate in the Electrochemical Science and Engineering group led by Professor Nigel Brandon, involved with the Energy Storage for Low Carbon Grids project. In 2017 he started his permanent position in the Dyson School of Design Engineering where he founded the TLDR group.
Sam's group have recently developed a method for using Generative AI to optimise the microstructure of advanced materials such as battery electrodes, alloys, and cements. The group have also published a method for generating 3D samples from 2D image data, which is vital for a variety of materials characterisation activities.
Sam and his team have developed several open-source packages that have gathered a diverse user base within the materials microstructure community, which can all be accessed through the group's github page.
A few well-cited examples that are applicable to a wide range of materials:
* TauFactor is a tool for the rapid analysis of 3D microstructural data derived for various tomography techniques: https://www.theoj.org/joss-papers/joss.05358/10.21105.joss.05358.pdf
* SliceGAN is a AI-based model for generating 3D microstructural data from a single 2D image: https://www.nature.com/articles/s42256-021-00322-1
* ImageRep is a method for calculating the representativity of a microstructural metric from a single image: www.imagerep.io
* SambaSegment is an image segmentation tool that makes use of Meta's Segment Anything Model: www.sambasegment.com
ACADEMIC POSITIONS
- ReaderImperial College London, Dyson School of Design Engineering, London, United Kingdom1 May 2017 - present
NON-ACADEMIC POSITIONS
- Cheif ScientistPolaron, London1 May 2023 - present
DEGREES
- MEngImperial College London, London, United Kingdom5 Oct 2008 - 7 Jul 2012
- PhDImperial College London, London, United Kingdom1 Oct 2012 - 6 Nov 2015
FACULTY
- Faculty of Engineering
POSITION NAME
- Associate Professor in Machine Learning for Materi
MEDIA GUIDE
- Members of the media are welcome to contact me about my research and areas of expertise