This project builds an automated geometry analysis pipeline using Keyence laser displacement / line-profile metrology as the in-line measurement backbone. The workflow consistently performs preprocessing, alignment, filtering, feature recognition, and parameterization from raw profiles, accounting for production-scale constraints such as noise, missing samples, and reference-frame variability. The goal is to convert raw metrology streams into quantitative, decision-ready geometry indicators that can be directly used for manufacturing quality evaluation.

The primary objective is to quantitatively characterize fuel-cell bipolar plate channel geometry and reliably extract geometric parameters for quality assessment and process-condition optimization. The target features include channel length, radii of curvature, channel depth (height), and sidewall angle. These parameters directly affect flow resistance, reactant gas distribution, and water management performance, and therefore play a critical role in the overall efficiency and durability of fuel cell stacks.

Beyond offline analysis, the project includes the development of a GUI-based interface that enables streamlined operation from measurement to reporting. The system automatically recognizes repeating channel features and provides standardized outputs—parameter tables and visual diagnostics—supporting deviation checks against design envelopes and traceable quality screening. This GUI-centered toolchain facilitates practical deployment in production environments by reducing operator burden and enabling consistent, repeatable geometry evaluation.