Trustworthy Traceability for Logistics Chains Using Latent Fingerprints and Quality-Aware Score Fusion

Abstract

Modern logistics chains involve multiple handover stages, during which evidence can be easily lost or tampered with, making responsibility tracing difficult. To address this issue, this paper proposes a quality-aware trustworthy evidence preservation and provenance method based on latent fingerprints. The method constructs an event-driven evidence package for each logistics event and adopts a hybrid storage architecture: event metadata and evidence digests are recorded on-chain, while latent fingerprint data and auxiliary materials are encrypted and stored off-chain, thereby forming a verifiable evidence chain. In addition, a quality-aware mechanism fuses latent fingerprint quality indicators with matching scores to improve evidence usability. We treat latent fingerprint usability as part of the provenance record by storing a quality-fused credibility score to support dispute investigation. Experimental results show that the proposed quality-fusion strategy outperforms a matcher-only baseline in verification performance. In the overhead evaluation, the scheme achieves higher write efficiency and lower on-chain storage footprint under typical workloads, meeting the requirements for continuous event recording and rapid trace-back in logistics operations. Moreover, the proposed method can effectively detect tampering attacks, thereby ensuring evidence integrity and enabling trustworthy traceability in logistics scenarios. Overall, the proposed approach provides a technical path that balances usability, trustworthiness, and compliance for logistics dispute evidence collection and responsibility attribution.