Reliable Quality Traceability for Tilapia Cold Chain Using Blockchain and Machine Learning Techniques

Tilapia are easily prone to degradation during the cold chain process, which is an urgent need for a transparent, efficient, and trustworthy traceability system. This paper designed and implemented a tilapia-blockchain IoT traceability system (T-BITS) based on Hyperledger Fabric. Intelligent sensing device and smart contracts were developed for traceability modeling and consensus optimization. Furthermore, a machine-learning approach was used to achieve quality grading evaluation for tilapia cold chain. The GWO-LSTM-based key parameters prediction and PSO-SVM-based quality grading model were established. The results show that the T-BITS system is more effective to capture and trace the critical ambient and quality information for tilapia cold chain. PSO-SVM model accuracy for quality coupling grading reaches 93.33%. This work can provide decision-making reference for tilapia quality control.