Intelligent Transportation Systems and Logistics (ITS&L) are rapidly evolving and have become a cornerstone of the modern socio‑economic landscape, relying heavily on massive data, artificial intelligence (AI), and machine learning (ML).
This paper provides a comprehensive review of trustworthy Data Operations (DataOps) and Machine Learning Operations (MLOps) within the ITS&L domain. It first identifies gaps in existing literature regarding practical pipelines, tool selection, and trustworthiness assessment, thereby defining a clear research niche.
Key components of DataOps include data ingestion, cleaning, governance, versioning, and real‑time monitoring. Popular open‑source frameworks such as Apache Airflow, Dagster, and Great Expectations enable workflow orchestration, data quality validation, and metadata management.
MLOps focuses on end‑to‑end model training, validation, deployment, monitoring, and rollback. Platforms like Kubeflow, MLflow, and TensorFlow Extended (TFX) provide experiment tracking, model registry, and scalable serving capabilities for both batch and streaming inference.
To strengthen AI trustworthiness, the paper highlights interpretability methods (SHAP, LIME), robustness testing, model verification, and compliance auditing, recommending tools such as WhyLabs and Evidently AI for visual monitoring of drift and anomalies.
Practical case studies cover city traffic flow forecasting, logistics route optimization, and fleet dispatching. These examples illustrate how DataOps and MLOps pipelines transform raw sensor streams into cleaned datasets, feed time‑series models, and finally deploy models on edge devices for real‑time decision making, markedly improving efficiency and reliability.
Nevertheless, challenges remain: cross‑domain data privacy, tension between real‑time performance and scalability, lack of standardization, and regulatory compliance.
Future research directions include adaptive pipelines, federated learning for multi‑party data collaboration, digital‑twin‑driven simulation testing, and holistic sustainability assessment of ITS&L systems.
Review