A Three-Layer Blockchain Framework for Smart Cold Chain Traceability of Fresh Agricultural Products

Ensuring end-to-end traceability in fresh agricultural product cold chains remains a persistent challenge, as existing systems suffer from centralised data vulnerability, absence of automated enforcement mechanisms, and disconnection between traceability records and analytical decision-support tools. This paper proposes a three-layer blockchain architecture based on Hyperledger Fabric. The system integrates smart contracts and big data analytics to strengthen traceability and supply chain transparency. It is designed to address operational inefficiencies in the distribution of temperature-sensitive perishable goods. The framework structures cold chain management into three layers: Data, Contract, and Application. The Data Layer adopts a hybrid on-chain/off-chain storage mechanism with hash anchoring to ensure integrity and scalability. The Contract Layer deploys three specialised smart contracts for quality certification, transaction settlement, and cross-organisational information sharing. The Application Layer integrates a big data analytics module that performs parameter optimisation and feeds the results back into the contract layer, forming a closed-loop system. Comparative analysis against four representative existing approaches demonstrates that the proposed architecture offers broader smart contract coverage and deeper analytical integration than current solutions. The closed-loop feedback mechanism transforms the traceability system from a static recording tool into an adaptive management platform. It is capable of continuous operational refinement. The contribution of this paper lies in the design and conceptual evaluation of the proposed architecture rather than in an empirically validated implementation, and prototype deployment together with quantitative performance assessment is planned as a subsequent research stage. Limitations concerning the dependency of the framework on Internet of Things (IoT) infrastructure, the scalability of transaction throughput under high-frequency monitoring conditions, and the governance of automated parameter updates are acknowledged as directions for further investigation.