ResinSpec

ResinSpec is a web app (with optional desktop uploader) that helps labs and small manufacturers identify unknown plastics and verify incoming resin lots using common analytical outputs like FTIR spectra, DSC curves, and TGA traces. Users upload instrument exports (CSV/JDX/PDF), the app cleans and standardizes the data, then matches it against a curated reference library of polymers, additives, and common blends. It produces a confidence-ranked identification, flags likely contaminants (e.g., PVC in PET stream), and generates a simple pass/fail report aligned to user-defined specs (melt peak range, glass transition, oxidation onset). This is a combination traditional + AI app: traditional rules/chemometrics for baseline reliability, plus ML for pattern matching and anomaly detection. It’s realistic because it targets organizations that already run these tests but waste time interpreting results and writing reports.

← Back to idea list