Data Integrity Validation For Used Cooking Oil Logistics Based On Haversine Algorithm And Image Metadata

Validasi Integritas Data untuk Logistik Minyak Goreng Bekas Berdasarkan Algoritma Haversine dan Metadata Gambar

Abstract

The implementation of Indonesia’s B50 biodiesel mandate faces weak data verification within the used cooking oil (UCO) reverse logistics chain managed by MSMEs. Conventional trackers cannot validate the authenticity of field coordinates and photographic evidence, creating data vulnerabilities. This study developed the Capture-Guard Framework, a low-cost solution combining the Haversine Algorithm for spatial validation and EXIF metadata parsing for temporal-device validation. An experimental simulation was executed using an expanded synthetic dataset of 200 comparative entries. By evaluating logs against a 10-meter spatial boundary and a 120-second temporal tolerance protocol, the framework successfully filtered all malicious logs, yielding an overall binary classification accuracy of 93.5%. The system achieved a perfect Precision rate of 100% due to zero False Positive occurrences, while environmental noise and sensor drifts were safely isolated under a tiered verification class, resulting in a Recall rate of 87.0% and a robust F1-Score of 0.93. These empirical findings demonstrate that securing data integrity at the first-mile validation point provides a clean audit trail vital for supporting raw material certainty within the B50 biodiesel ecosystem