A

Deep learning for decoding lithium plating in lithium-ion batteries from post-mortem images

Journal of Energy Storage, vol. 154, pp. 121028

Abstract

Lithium plating is a safety-critical parasitic process in lithium-ion batteries that reduces efficiency and usable capacity and elevates safety risk. Conventional assessment relies on manual post-mortem inspection, but manual inspection is slow, subjective, and hard to scale. This work introduces an automated computer-vision methodology for the systematic characterization of lithium plating. A convolutional attention-based variational autoencoder with k -means clustering identifies and quantifies plating features from high-resolution flatbed scans. The model was trained on the complete imaging set (over 5000 scans from more than 60 cells) to construct a descriptive morphological atlas of the experimental matrix. A fixed condition-complete subset of 13 cells ( N = 1 , 092 scans) was used solely as a convergence monitor and for the quantitative trend analyses reported in this work. The learned two-dimensional latent space disentangles C-rate (1C to 4C) and temperature ( − 2 0 ∘ C to − 5 ∘ C ) while tracking plating severity and spatial uniformity; its coordinates correlate monotonically with the stripping capacity Q strip , providing a quantitative bridge between surface morphology and electrochemical indicators. Spearman analysis confirms these latent axes encode operating conditions: Dim1 vs Temp | ρ | =0.93, Dim2 vs Temp | ρ | =0.84, Dim1 vs C-rate | ρ | =0.75, Dim2 vs C-rate | ρ | =0.72. The approach delivers objective and scalable diagnostics that complement non-invasive electrical methods and demonstrates that computer vision can capture operating-condition effects directly from post-mortem images, enabling risk maps and high-throughput screening to support safer and longer-lasting lithium-ion battery systems.

Authors 6

  1. Jülich Aachen Research Alliance · RWTH Aachen University

    Affiliation as printed

    Center for Aging, Reliability and Lifetime Prediction of Electrochemical and Power Electronic Systems (CARL), RWTH Aachen University, Campus-Boulevard 89, Aachen, 52074, Germany

    Institute for Power Electronics and Electrical Drives (ISEA), RWTH Aachen University, Campus-Boulevard 89, Aachen, 52074, Germany

    Juelich Aachen Research Alliance, JARA-Energy, Templergraben 55, Aachen, 52056, Germany

  2. Jülich Aachen Research Alliance · RWTH Aachen University

    Affiliation as printed

    Center for Aging, Reliability and Lifetime Prediction of Electrochemical and Power Electronic Systems (CARL), RWTH Aachen University, Campus-Boulevard 89, Aachen, 52074, Germany

    Institute for Power Electronics and Electrical Drives (ISEA), RWTH Aachen University, Campus-Boulevard 89, Aachen, 52074, Germany

    Juelich Aachen Research Alliance, JARA-Energy, Templergraben 55, Aachen, 52056, Germany

  3. Jülich Aachen Research Alliance · RWTH Aachen University

    Affiliation as printed

    Center for Aging, Reliability and Lifetime Prediction of Electrochemical and Power Electronic Systems (CARL), RWTH Aachen University, Campus-Boulevard 89, Aachen, 52074, Germany

    Institute for Power Electronics and Electrical Drives (ISEA), RWTH Aachen University, Campus-Boulevard 89, Aachen, 52074, Germany

    Juelich Aachen Research Alliance, JARA-Energy, Templergraben 55, Aachen, 52056, Germany

  4. Jülich Aachen Research Alliance · RWTH Aachen University

    Affiliation as printed

    Center for Aging, Reliability and Lifetime Prediction of Electrochemical and Power Electronic Systems (CARL), RWTH Aachen University, Campus-Boulevard 89, Aachen, 52074, Germany

    Institute for Power Electronics and Electrical Drives (ISEA), RWTH Aachen University, Campus-Boulevard 89, Aachen, 52074, Germany

    Juelich Aachen Research Alliance, JARA-Energy, Templergraben 55, Aachen, 52056, Germany

  5. Helmholtz-Institute Münster · Jülich Aachen Research Alliance · RWTH Aachen University

    Affiliation as printed

    Center for Aging, Reliability and Lifetime Prediction of Electrochemical and Power Electronic Systems (CARL), RWTH Aachen University, Campus-Boulevard 89, Aachen, 52074, Germany

    Helmholtz Institute Muenster (HI MS), IMD-4, Forschungszentrum, Juelich, Germany

    Institute for Power Electronics and Electrical Drives (ISEA), RWTH Aachen University, Campus-Boulevard 89, Aachen, 52074, Germany

    Juelich Aachen Research Alliance, JARA-Energy, Templergraben 55, Aachen, 52056, Germany

  6. Jülich Aachen Research Alliance · RWTH Aachen University

    Affiliation as printed

    Center for Aging, Reliability and Lifetime Prediction of Electrochemical and Power Electronic Systems (CARL), RWTH Aachen University, Campus-Boulevard 89, Aachen, 52074, Germany

    Institute for Power Electronics and Electrical Drives (ISEA), RWTH Aachen University, Campus-Boulevard 89, Aachen, 52074, Germany

    Juelich Aachen Research Alliance, JARA-Energy, Templergraben 55, Aachen, 52056, Germany

Cited by 0 stored of 0

No patents citing this paper on Lens.org (checked 2026-10-06).

References 32