EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A Multi-Modal Framework for Detecting Physical–Digital Container Identity Drift in Global Logistics

Manikandan Chandran

Engineering, Technology & Applied Science Research · 11.08.2026 · 16 (4) · 37657-37666

DOI: 10.48084/etasr.19710 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Engineering, Technology & Applied Science Research, online erschienen am 11.08.2026. Verfasst von Manikandan Chandran. Die Arbeit nennt 24 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This study addresses the problem of physical–digital container identity drift in global logistics, with the explicit objective of designing and evaluating a multi-modal, edge-deployable framework that detects sustained inconsistencies between a container's observable physical identifiers and its associated digital records in near real time. Modern container logistics relies on the continuous alignment of physical identifiers and digital records, yet this alignment frequently degrades due to repainting, reuse, environmental factors, and operational inconsistencies. Existing approaches rely on single-modality verification such as Optical Character Recognition (OCR) or document validation, which are insufficient in complex port environments. This paper proposes a cross-modal consistency framework deployed at the edge that integrates container imagery, OCR-extracted identifiers, Automatic Identification System (AIS)-derived vessel context, geospatial trajectories, and logistics metadata. A consistency-driven fusion engine evaluates pairwise agreement across modalities through six closed-form similarity functions, and a rule-augmented anomaly-scoring model identifies drift events. To address limitations of prior work, a hybrid evaluation dataset is constructed combining real-world container imagery with controlled perturbations representing visual, spatial, document, and temporal drift scenarios. Experiments executed on an NVIDIA Jetson Nano edge platform show that the proposed framework improves visual-drift F1 from 0.66 to 0.86, spatial-drift F1 from 0.59 to 0.84, document-drift F1 to 0.88, and matches the strongest learned baseline on temporal drift (F1 = 0.83), while sustaining 120 ms average end-to-end latency, 30 events/s throughput, and a 120 KB per-event bandwidth footprint. The proposed approach reframes container identity validation as a cross-modal consistency problem and provides a practical, deployable solution for enhancing supply-chain security and operational reliability.

Zitieren

APA

Chandran, M. (2026). A Multi-Modal Framework for Detecting Physical–Digital Container Identity Drift in Global Logistics. Engineering, Technology & Applied Science Research, 16(4), 37657-37666. https://doi.org/10.48084/etasr.19710

Vancouver

Chandran M. A Multi-Modal Framework for Detecting Physical–Digital Container Identity Drift in Global Logistics. Engineering, Technology & Applied Science Research. 2026;16(4):37657-37666. doi:10.48084/etasr.19710

BibTeX
@article{Chandran2026,
  author = {Chandran, Manikandan},
  title = {A Multi-Modal Framework for Detecting Physical–Digital Container Identity Drift in Global Logistics},
  journal = {Engineering, Technology & Applied Science Research},
  year = {2026},
  volume = {16},
  number = {4},
  pages = {37657--37666},
  doi = {10.48084/etasr.19710}
}
Kostenlos starten →