Industries›EV Manufacturing

Vision AI for
EV Manufacturing

Ensuring the quality assurance and stability of EV batteries is a key competitive factor for manufacturers. Therefore, 3D In-Line inspection techniques, which inspect the internal components of products during the manufacturing process, are being introduced.

In the field of 3D In-Line inspection, deep learning-based data processing techniques such as denoising and Super-Resolution (SR), as well as models for detecting or classifying internal defects within battery cells, are being applied.

Inquiry

EV Manufacturing

Use Cases

EV Battery Cell Defect(Segmentation) and Post Processing model for assessing the defect scale

Issues and
Needs to Address

  • A decreased Takt Time for data collection in In-Line inspection processes results in lower data quality.
  • The technology required to classify various types of defects in battery packs or track the location of defects is essential.
  • The technology required for quantitatively analyzing the degree of bending of electrodes is crucial for assessing the extent of defects in batteries.

Solutions and
Applied Technologies

  • Applied SNUAILAB Image Restoration Network (SIRNet) to resolve the issue of decreased Takt Time.
  • Applied SNUAILAB Inspection Network (SINNet) to support the identification of defects and their location in images for battery cell data training.
  • Developed a Post Processing Model for quantitatively analyzing the degree of bending inside battery cells.

Results of Technology Implementation

  • SIRNet — before (original) / after restoration
  • SINNet + post-processing — semantic segmentation / segmentation + post-processing

Are you curious about SNUAILAB's AI technology
applied across various industries?

Maximize your business values
by adopting industry-specific AI solutions.

Inquiry