Dr Ying Da Wang, Dr Quentin Meyer, Prof Ryan Armstrong and Prof Chuan Zhao and their multidisciplinary team developed an algorithm to improve images of hydrogen fuel cells, with future applications in medical scanning. We sat down with Dr Quentin Meyer for more information.
How the project came to be.
Hydrogen fuel cells use hydrogen to generate electricity and are a quiet and clean energy source that can power homes, vehicles and industries. Via an electrochemical process, fuel cells convert hydrogen into electricity with the only by-product of the reaction being pure water.
However, hydrogen fuel cells become inefficient if the water is not removed fast enough and floods the system. Our team developed the algorithm to understand how this flooding happens inside a fuel cell.
Figure 1. 3D X-ray scan of a hydrogen fuel cell, showing carbon paper weaves, membrane, and catalysts (in black). Image provided by Dr Quentin Meyer
One of my collaborators from Zeiss had some interesting images at different resolutions and after discussing it with Prof Ryan Armstrong [of the School of Minerals and Energy Resources Engineering, University of New South Wales] one year ago, we quickly got very excited.
The project eventually grew to a large team, with 13 researchers over three continents in Australia, the UK and the USA, three research institutes and two companies, with expertise in different areas from fuel cell engineering and advanced imaging to fluid transport in porous media.
We developed an algorithm to produce high-resolution modelled images from low-resolution micro-X-ray computed tomography (CT).
This exciting new tool, as published in Nature Communications, was evaluated on hydrogen fuel cells to model their structure with higher details and potentially improve their efficiency.
DualEDSR algorithm.
Our algorithm uses deep learning to create a detailed 3D model, using a low-resolution X-ray image of the cell and extrapolating data from an accompanying high-res scan of a small sub-section.
It’s like taking a blurry aerial view of a town, a photo of a few streets, and then accurately predicting the layout of every road in the area. This is particularly exciting because we are pushing the imaging resolution limits by a few orders of magnitude using machine learning and artificial intelligence [AI].
Our DualEDSR algorithm improves the field of view by around 100 times compared to the high-res image. If you look at what we are doing now and apply it to the medical field, we could, for instance, image blood vessels and model the flow of red blood cells through the capillary network in detail.
These beyond-hardware imaging and modelling methods extend beyond fuel cell imaging to enable higher-resolution imaging of larger fields of view than anything previously possible.
Currently, the large-scale, low-resolution image and the small-scale, high-resolution image need to be at the same location on the same machine, which creates some limitations.
However, future research should soon allow deep learning techniques to produce similar results without matching areas or even instruments. During training and testing, the algorithm was 97.3% accurate in producing high-resolution modelling from low-resolution images.
Water transport in PEMFCs.
It produced a high-resolution model in one hour, compared to the 1,188 hours needed to scan the whole section of the fuel cell at similar resolution using a micro-CT scanner.
From our model, we can quickly and precisely identify where water accumulates and potentially solve problems in future designs. There is a huge untapped performance improvement that could be made using these cells, just by improved water management—estimated to be 60% overall.
For the past 20 years, it has been very hard to have an accurate model of these fuel cells because of the complexity of the materials and the way gases and liquids are transported, as well as the electrochemical reactions taking place.
Our team has done just that, bringing different expertise to the table. In my opinion, this is what research is about.