The system and methodology for evaluating the quality, classification, and traceability of durian uses artificial intelligence with multiple data sources combined with verifiable digital quality records.

Abstract

This invention proposes a system and method for evaluating the quality, classification, and traceability of durian using multi-source artificial intelligence combined with verifiable quality records. The system receives image data, audio feedback from typing operations, weight, size, VOC signals, ethylene signals, growing area data, harvesting data, packaging data, logistics data, and post-inspection data; standardizes the data; and assigns an identifier to each fruit, carton, pallet, or batch of durian. The data is analyzed by AI to assess appearance, ripeness level, risk of defects, quality score, classification status, and reliability score. The system controls the classification mechanism, creates durian quality records, generates evidence packages, records hash codes and timestamps, and references data to a registry or blockchain for verification, traceability, post-inspection, and AI model retraining.