% IMPORTANT: The following is UTF-8 encoded. This means that in the presence
% of non-ASCII characters, it will not work with BibTeX 0.99 or older.
% Instead, you should use an up-to-date BibTeX implementation like “bibtex8” or
% “biber”.
@INPROCEEDINGS{Colliard:1038801,
author = {Colliard, Andre and Malek, Kourosh and Eikerling, Michael
and Eslamibidgoli, Mohammad Javad},
title = {{D}eep {L}earning-driven {A}utonomous {I}maging
{C}haracterization of {P}olymer {E}lectrolyte {M}embrane
{F}uel {C}ell {T}echnologies},
reportid = {FZJ-2025-01641},
year = {2024},
abstract = {Seamless integration of deep learning and computer
vision-driven tools has become essential to accelerate
research addressing global energy challenges. This is
particularly true in the field of energy technology
characterization, where the need for high throughput and
advanced methods creates a bottleneck in the analysis and
interpretation of large data volumes. In this study, we
introduce three different use-cases related to autonomous
image analysis of polymer electrolyte membrane technologies
across nano-, micro-, and macro-scales, aimed at
accelerating and enhancing insights in this domain. For each
case, a comprehensive tool was developed, integrating image
preprocessing, deep learning model predictions, feature
extraction, and results visualization, thereby facilitating
usage by experimentalists.At the nanoscale, we have
developed a tool for autonomous 2D Transmission Electron
Microscopy (TEM) image analysis to assess the size and shape
of Platinum nanoparticles supported on Carbon. aimed at
investigating the catalyst layer of polymer electrolyte
membrane fuel cells (PEMFC).[1] Progressing to the
microscale, we introduce a deep learning-based approach for
analyzing micro-CT 3D volumes, focusing on pore analysis.
This software is designed to perform multi-class
segmentation of pores and liquid channels within gas
diffusion layers (GDL), providing detailed insights into
pore structures and water transport properties.[2] Lastly,
at the macroscale, we showcase a tool developed for
detecting oxygen bubbles in optical videos from polymer
electrolyte water electrolyzers (PEMWE). This application
enables the time-resolved analysis of bubble dynamics within
the water electrolyzer’s flow field, offering a
macroscopic view of gas evolution and its impact on device
performance.[3] Through these tools, our work demonstrates
the potential of deep learning and computer vision to bridge
the gap between scales in energy materials research,
offering novel insights and expediting the analysis
process.[1] André Colliard-Granero et al., “Deep
learning for the automation of particle analysis in catalyst
layers for polymer electrolyte fuel cells,” vol. 14, no.
1, pp. 10–18. [2] André Colliard-Granero et al., “Deep
Learning for Automation of 3D Pore Analysis in Micro-CT
Tomographs,” under preparation. [3] André
Colliard-Granero et al., “Deep Learning-Enhanced
Characterization of Bubble Dynamics in Proton Exchange
Membrane Water Electrolyzers,” Physical Chemistry Chemical
Physics, 2024, Accepted Manuscript. DOI:
https://doi.org/10.1039/D3CP05869G},
month = {Oct},
date = {2024-10-06},
organization = {Pacific Rim international meeting
2024, Honolulu (USA), 6 Oct 2024 - 11
Oct 2024},
subtyp = {After Call},
cin = {IET-3},
cid = {I:(DE-Juel1)IET-3-20190226},
pnm = {1231 - Electrochemistry for Hydrogen (POF4-123)},
pid = {G:(DE-HGF)POF4-1231},
typ = {PUB:(DE-HGF)6},
url = {https://juser.fz-juelich.de/record/1038801},
}