Trustworthy Data- and ML-Ops for Intelligent Transportation Systems and Logistics

Antonio Emanuele Cinà; Giovanni Scodeller; Cecilia Caterina Pasquale; Silvia Siri; Davide Anguita; Fabio Roli; Simona Sacone; Luca Oneto

cs.AIOct 1, 2026

Trustworthy Data- and ML-Ops for Intelligent Transportation Systems and Logistics

Authors: Antonio Emanuele Cinà, Giovanni Scodeller, Cecilia Caterina Pasquale, Silvia Siri, Davide Anguita, Fabio Roli, Simona Sacone, Luca Oneto

Abstract

The rapid evolution of Intelligent Transportation Systems and Logistics (ITS&L) has become a cornerstone of the modern social economy, relying heavily on the integration of Data, Artificial Intelligence (AI), and, more specifically, Machine Learning (ML). This paper provides a comprehensive review of Trustworthy Data and Machine Learning Operations (DataOps and MLOps) in the ITS&L domain, underscoring their importance in improving efficiency, reliability, and decision-making precision within transportation and logistics services. We begin by identifying gaps in current literature, offering clear context for our contribution. Subsequently, we explore the complexities of DataOps and MLOps, discussing their necessity, key components, available tools, practical insights, and case studies relevant to ITS&L. Additionally, we address the critical issue of Trustworthiness in AI applications, examining methods and tools designed to strengthen confidence in AI systems - especially in real-world ITS&L scenarios. The paper concludes with a discussion of persisting challenges and future prospects in this rapidly advancing field, aiming to serve as a vital resource for researchers, industry practitioners, and policy makers. Overall, this work not only establishes a foundational understanding of DataOps and MLOps in ITS&L but also charts a path for further research and innovation in developing more efficient, sustainable, and trustworthy intelligent transportation and logistics systems.

Figures & tables

Figure 1: DataOps and MLOps in ITS&L. Heterogeneous data sources are governed through the DataOps lifecycle and subsequently support the MLOps lifecycle for model development, deployment, monitoring, and updating. The resulting AI-enabled services operate across the main ITS&L application domains, while TAI provides a governance layer grounded in robustness, explainability, fairness, privacy, and sustainability.

Ref.	Year	ITS&L	DataOps	MLOps	TAI	Keywords
TrustworthyAI Surveys
[ 49 ]	2018					Adversarial machine learning, Evasion attacks, Poisoning attacks, Adversarial examples, Secure learning, Deep learning
[ 50 ]	2021					Trustworthy AI, robustness, generalization, explainability, transparency, reproducibility, fairness, privacy protection, accountability
[ 28 ]	2021					Machine learning, privacy, deep learning, differential privacy
[ 51 ]	2022					Learning on graphs, Graph neural networks, Kernels for graphs, Trustworthy machine learning, Fairness, Privacy, Robustness, Explainability, Learning automatically Learning with guarantees
[ 26 ]	2023					Poisoning attacks, backdoor attacks, Machine learning, Computer vision, Computer security

TABLE I: Meta-review of existing surveys in the ITS&L, DataOps , MLOps , and TAI research fields.

Figure 2: DataOps abstract lifecycle.

Open-source – core data processing tools
Tool	Source	DataOps phase(s)	Processing	On-premise	Scalability	Limitation
Apache Kafka	Open-source	Ingestion, Monitoring	Stream		High throughput and low latency for large-scale streams	Requires expertise to configure, operate, and maintain distributed streaming clusters
Apache HDFS	Open-source	Storage	–		Horizontal scalability through data replication and node addition	Primarily designed for storage and batch-oriented workloads, requiring complementary tools for real-time analytics
Apache Spark	Open-source	Processing	Batch		Distributed in-memory processing across multiple nodes, reducing disk access	Cluster configuration and performance tuning can be complex in production environments
Apache Flink	Open-source	Processing	Stream		Reduces storage overhead by processing continuous streams	Operationally more complex than batch-oriented frameworks and often requires strong stream-processing expertise
Open-source – Ops tools, from release to feedback

TABLE II: Comparison of representative DataOps tools.

Figure 3: MLOps abstract lifecycle.

Open-source – core ML pipeline tools
Tool	MLOps phase(s)	AutoML	On-premise	Scalability	Limitation
Metaflow	Data engineering, training, validation, deployment	×		On-demand cloud integration for large-scale workloads	Does not cover the full MLOps pipeline; monitoring and feedback stages require additional tools
ZenML	End-to-end	×		Via third-party integrations with cloud and on-premise services	Relies on external tools, such as MLflow, Weights and Biases, and BentoML, to cover several pipeline stages, adding integration overhead
BentoML	Release, packaging, deployment, serving, monitoring support	×		Docker, Kubernetes, and cloud services via Bentoctl	Focuses mainly on release, deployment, and serving rather than on the full training and validation lifecycle
Kubeflow	End-to-end			Distributed execution across Kubernetes clusters	Tightly coupled to Kubernetes infrastructure, requiring significant DevOps expertise to deploy and maintain
Apollo	Data engineering, validation, deployment, monitoring	×		Simulation-based testing environments for vehicle control applications	Not designed for general ML workflows domains beyond vehicle control

TABLE III: Comparison of representative MLOps tools.

Areas	Ref(s)	# Ref(s)	DataOps	MLOps	TAI
Areas	Ref(s)	# Ref(s)	DataOps	MLOps	Sustain.	Robust.	XAI	Fairness	Privacy
Vehicle Control	[ 122 , 180 , 178 , 179 , 180 , 181 , 182 , 183 , 184 , 185 , 186 , 187 ]	12
	[ 280 , 281 , 282 ]	3
	[ 493 , 494 , 495 , 496 , 497 , 498 , 499 , 500 , 501 ]	9
	[ 489 , 490 , 491 , 502 , 492 ]	5
	[ 484 , 486 , 487 , 488 ]	4

TABLE IV: Case studies summary for DataOps , MLOps , and TAI .

Explore similar work

Large Language Models in Transportation Systems Management and Operations: From Text Reasoning to Multi-modal Decision Support
May 31, 2026Siyan Li, Zehao Wang, Jiachen Li +3Multimodal Large Language Models Software Engineering Decision Support
Bourns College of Engineering, Center for Environmental Research and Technology, University of California at Riverside, CA, USA · Bourns College of Engineering, Computer Science and Engineering Department, University of California at Riverside, CA, USA · Bourns College of Engineering, Electrical and Computer Engineering Department, University of California at Riverside, CA, USA
Large Multimodal Agents for Intelligent Transportation Systems: Architectures, Evidence, and Deployment Challenges
Aug 8, 2026Muhammad Ayub Sabir, Shaohong Zheng, Zhiyu Qu +2Multimodal Agents Model-Based Multi-Agent Systems
A Comparative Analysis of Machine Learning Models for Intrusion Detection in Intelligent Transport Systems
Apr 30, 2026Zawad Yalmie Sazid, Robert Abbas, Sasa MaricEdge Devices Intrusion Detection
Victoria University, Sydney, Australia · University of New South Wales College, Sydney, Australia

May 31, 2026, cs.AIJ/K move · Enter open · S save

Large Language Models in Transportation Systems Management and Operations: From Text Reasoning to Multi-modal Decision Support

Siyan Li, Zehao Wang, Jiachen Li, Kanok Boriboonsomsin+2

Bourns College of Engineering, Center for Environmental Research and Technology, University of California at Riverside, CA, USA · Bourns College of Engineering, Computer Science and Engineering Department, University of California at Riverside, CA, USA · Bourns College of Engineering, Electrical and Computer Engineering Department, University of California at Riverside, CA, USA

Multimodal Large Language Models Software Engineering Decision Support Transport Reasoning Skills

Open paper AskPDF arXiv

Transportation systems management and operations (TSMO) increasingly depends on timely interpretation of heterogeneous data, from various sensor streams, incident reports, traveler feedback, and visual observations. Large language models (LLMs), including emerging multi-modal large language models (MM-LLMs), provide a new mechanism for integrating these structured and unstructured inputs into operator-facing decision support. This survey paper reviews LLM- and MM-LLM-based applications in TSMO across three domains: transportation operations & services (supply), mobility & fleet services (demand), and data, modeling & decision support. Using a PRISMA-guided screening process, we synthesize current studies while distinguishing operationally oriented applications from prototype and emerging concepts. We further identify recurring challenges in data heterogeneity, real-time inference, explainability, multi-modal fusion, and governance. Finally, we outline existing gaps and future directions in localized adaptation, edge deployment, benchmarking, and cross-agency collaboration. Overall, LLM-based systems appear most promising as a decision-support layer, with MM-LLMs offering particular value when heterogeneous text, visual, and sensor inputs must be integrated.

Abstract

Explore similar work

Large Language Models in Transportation Systems Management and Operations: From Text Reasoning to Multi-modal Decision Support

Large Multimodal Agents for Intelligent Transportation Systems: Architectures, Evidence, and Deployment Challenges

A Comparative Analysis of Machine Learning Models for Intrusion Detection in Intelligent Transport Systems