Calls for Papers

Transportation Science Special Issue

Railway Challenges, Reproducibility, and Computational Breakthroughs

Submission Deadline: September 30, 2027; earlier submissions are encouraged

Description

This special issue seeks to catalyze the next wave of advances in railway operations research by connecting three essential elements:

• Real-World Challenges: Hard problems from actual railway operations worldwide
• Reproducibility: Open datasets, transparent methodologies, and verifiable results that enable independent validation and fair comparison across approaches
• Computational Breakthroughs: Methodological innovations—whether from classical operations research, modern optimization, machine learning, or hybrid approaches—that advance the state of the art

We invite the research community to revisit historical RAS competition problems with modern methods, to introduce new challenges from contemporary railway operations, and to demonstrate reproducible computational solutions that can impact practice. Greater details about the types of submissions expected and nature of problems considered are below. This special issue is organized in collaboration with the INFORMS Railway Applications Section (RAS) community, including Marty Schlenker, RAS Vice Chair 2025.

Submission Guidelines

Submissions will be peer-reviewed per the criteria listed below and in accordance with the standards of the journal. Please submit your manuscript online via ScholarOne ManuscriptCentral at http://mc.manuscriptcentral.com/transci.

When choosing the Manuscript Type in Step 1 of the submission procedure, choose the option “Special Issue: Railway Challenges, Reproducibility, and Computational Breakthroughs” to ensure your paper is considered for this special issue. Authors must specify their submission type (Type A or Type B, see descriptions below) in the cover letter. For questions about challenge problem suitability, dataset preparation, or submission process, please contact Professor Xuesong (Simon) Zhou at [email protected].

Timeline and Process

• Deadline for submissions: September 30, 2027; earlier submissions are encouraged
• Two rounds of review, completed on a rolling basis as papers are submitted to meet the deadline for final decision
• Spring 2028 (papers will be published online sooner, as per typical journal processes)

Guest Editors

• Lead: Xuesong (Simon) Zhou, Arizona State University, RAS Chair 2025 ([email protected])
• Christian Liebchen, Technische Hochschule Wildau
• Nikola Bešinović, Technical University of Dresden, RAS Chair 2023
• Shuai Su, Beijing Jiaotong University, RAS Chair 2024
• Marcella Samà, Roma Tre University, RAS Student Paper Competition Chair 2025
• Niels Lindner, Zuse Institute Berlin
• Yili (Kelly) Tang, University of Western Ontario, RAS Chair 2026

Motivation

Railway operations research has a distinguished history of solving large-scale, complex problems driven by real operational needs. In the 1990s, researchers developed column generation, very large-scale neighborhood search, and other breakthroughs that enabled United States Class I railroads to solve network routing and scheduling problems at unprecedented scale. European researchers pioneered cyclic timetabling methods for integrated passenger-freight networks. Asian railway systems have pushed the boundaries of high-frequency, precision operations and large-scale scheduling.

The INFORMS Railway Applications Section (RAS) has, over 15 years of problem-solving competitions (Data and Archive of Previous Problem Solving Competition - Railway Applications Section), accumulated a rich collection of real-world challenges spanning the full spectrum of railway planning and operations—from strategic network design to real-time dispatching, from predictive analytics to optimization under uncertainty. While there is no comprehensive collection of railway planning problems akin to established optimization libraries, rail OR and the broader research community would benefit greatly from a curated collection of challenging and practical instances representing the diverse tasks frequently encountered in freight and passenger railway systems.

This special issue seeks to catalyze the next wave of advances by connecting three essential elements:

• Real-World Challenges: Hard problems from actual railway operations worldwide
• Reproducibility: Open datasets, transparent methodologies, and verifiable results that enable independent validation and fair comparison across approaches
• Computational Breakthroughs: Methodological innovations—whether from classical operations research, modern optimization, machine learning, or hybrid approaches—that advance the state of the art

We invite the research community to revisit historical RAS competition problems with modern methods, to introduce new challenges from contemporary railway operations, and to demonstrate reproducible computational solutions that can impact practice. We particularly welcome problems from diverse railway systems worldwide, including, but not limited to:

• Freight-dominant systems
• Mixed passenger-freight networks
• High-speed and high-density operations
• Rapidly expanding metro systems
• Freight corridors in emerging markets
• Battery-powered local train operations and green transition initiatives

This special issue aligns with Transportation Science’s commitment to practice-inspired research and meaningful connections between methodological innovation and real-world impact (Hewitt, 2025). As emphasized in the Editor-in-Chief’s recent editorial statement, the journal values research that addresses well-motivated problems and demonstrates practical relevance alongside scientific rigor.

Note on Open Science: Open science is not the same as open data. We fully understand the need for data protection and confidentiality in industrial contexts. However, the RAS community already supports many well-defined benchmark problems that can form the basis for shared, verifiable research and education. By recognizing and contributing to open-science projects, industry and academia together can reproduce, apply, and advance research outcomes across multimodal and energy-integrated systems.

Scope and Topics

Building on 15 years of RAS competition problems and extending to new operational challenges, the special issue welcomes contributions on topics including, but not limited to, the following areas:

Network Planning and Operations

• Network routing and blocking (e.g., RAS 2019 and RAS 2026: Blocking, Path and Network Optimization)
• Service network design and capacity planning
• Train timetabling: cyclic and non-cyclic (e.g., RAS 2011, 2016, 2023: Train Design, Train Routing, Metro Timetabling)
• Rolling stock and crew scheduling (e.g., RAS 2010: Locomotive Refueling)
• Infrastructure investment and capacity expansion
• Integrated line planning, rolling stock, and crew planning

Real-Time Operations and Control

• Train dispatching and conflict resolution (e.g., RAS 2012, 2022: Train Dispatching, Real-Time Traffic Management)
• Delay management and robust operations
• Energy-efficient train control and trajectory optimization
• Dynamic resource allocation

Yard and Terminal Operations

• Classification yard operations (hump yards, e.g., RAS 2013, flat yards, e.g., RAS 2024)
• Block-to-track assignment (e.g., RAS 2014)
• Terminal capacity optimization
• Intermodal operations and transshipment

Data-Driven and Predictive Analytics

• Travel time and delay prediction (e.g., RAS 2020, 2021: Travel-Time Estimation, Predictive ETA)
• Demand forecasting and peak load prediction (e.g., RAS 2017)
• Railway system (e.g., vehicles, track, geometry) maintenance and planning (e.g., RAS 2015, RAS 2025)
• Real-time system state estimation

Integrated, Multimodal, and Sustainability Problems

• Passenger-freight network coordination
• Multi-operator systems on shared infrastructure
• Intermodal freight transportation
• Urban rail and metro operations
• Battery-powered trains, charging infrastructure, and rolling stock planning with charging
• Energy transition and sustainability in railway systems

Types of Submissions

The special issue invites two types of submissions. All submissions are expected to meet the cross-cutting reproducibility standards described in the next section. Both types require public data, but an RAS competition dataset is not required for either Type A or Type B.

A. Challenge Problem Papers

Present a real operational challenge with an accompanying open dataset. Papers should describe:

• The problem from the operator’s perspective and current practice
• Why the problem is computationally challenging
• Scale and complexity metrics
• Dataset description and access information (with a persistent identifier such as DOI, Zenodo, GitHub, or a stable repository link)
• Success criteria for solutions
• Potential operational impact of better solutions

Industry practitioners are strongly encouraged to co-author with academic researchers. We welcome contributions that revisit problems from the 15-year RAS competition history, applying modern methods while creating new, expanded, or enhanced datasets for open benchmarking.

B. Solution Method Papers

Demonstrate methodological advances on well-defined railway problems. Solution method papers may address historical RAS competition problems, newly introduced challenge problems (including those submitted as Type A to this special issue), or other openly available and well-documented railway challenge instances aligned with the special issue themes. Appropriate methodologies include, but are not limited to:

• Classical operations research (column generation, decomposition, branch-and-price, Lagrangian relaxation)
• Modern optimization (conic programming, constraint programming, network algorithms)
• Machine learning (deep learning, reinforcement learning, graph neural networks)
• Hybrid approaches (learning-enhanced optimization, predict-then-optimize, physics-informed ML)
• Simulation-based methods

Required elements:

• Solutions to specific, well-defined challenge problems using openly accessible datasets
• Reproducible methodology with computational environment details (hardware, software, solver versions)
• Performance comparison to baseline methods (when available)
• Code or software availability (e.g., link to a GitHub repository)
• Discussion of implementation context and practical relevance, when applicable

Note on timing: Type B submissions referencing a Type A challenge problem that is still under review are permitted, provided the problem instance and data are independently and publicly accessible.

Open Science and Reproducibility Standards

All submissions are expected to meet minimum reproducibility standards as a cross-cutting requirement of this special issue. Reproducibility expectations are framed as author responsibilities and evaluation criteria, and do not entail a separate journal-run verification process.

Minimum Expectations (required for all submissions)

• Problem instances publicly accessible with a persistent identifier (DOI, Zenodo, GitHub, etc.)
• Clear data format documentation
• Computational environment specification (hardware, software versions, solvers)
• Reporting of objective values, solution quality gaps, and computation times
• Random seeds and replication protocols for stochastic methods

Strongly Encouraged (recognized as additional contributions)

• Complete source code on a stable platform (e.g., GitHub, GitLab)
• Additional public datasets beyond the required minimum
• Containerized or scripted environments for experiment replication
• Baseline algorithm implementations for comparison
• Scripts for reproducing key experiments
• Supplementary documentation and tutorial materials

For papers making strong reproducibility claims, the guest editorial team will encourage authors to provide repositories, scripts, or supplementary materials that facilitate editorial and referee assessment. At least two guest editors will review computational reproducibility aspects as part of the standard editorial workflow.

Authors of accepted papers will be expected to upload supporting reproducibility materials to the INFORMS Transportation Science GitHub repository.

The materials and links of the special issue publications will be concurrently listed on the INFORMS Railway Applications Section website for promotion and reproducibility.

Evaluation Criteria

Papers will be evaluated on problem significance, methodological rigor, computational results, reproducibility, and practical relevance. Following Transportation Science guidelines (Hewitt, 2025), papers need not be outstanding in all dimensions but should excel in at least one. Evaluation emphasis varies by submission type:

Type A (Challenge Problem Papers): Emphasis on problem significance and real-world motivation, dataset quality and accessibility, clarity of formulation, and potential to stimulate future research. Papers from practitioner teams that include discussion of operational deployment experience are particularly valued.

Type B (Solution Method Papers): Emphasis on methodological contribution and rigor, computational performance relative to baselines, reproducibility of results, and scalability analysis. Discussion of implementation context and practical relevance is welcomed where applicable.

Submission-type-specific referee guidelines will be developed in consultation with the Area Editors and Editor-in-Chief, building on the standard Transportation Science referee guidelines.

Submissions will follow the standard Transportation Science referee guidelines, with the submission-type-specific evaluation emphases described above.

Why Submit to This Special Issue?

• Real Impact: Address problems that matter to railway operators worldwide
• Fair Comparison: Common datasets enable rigorous method evaluation
• Reproducibility: Enhanced visibility and credibility through transparent research
• Community Building: Connect with a global network of railway operations researchers
• Practice-Inspired Research: Aligns with Transportation Science’s emphasis on practically relevant, impactful research (Hewitt, 2025)

Building on the legacy of column generation breakthroughs in the 1990s and the 15-year RAS competition tradition, this special issue aims to establish new standards for problem-driven, reproducible railway operations research that bridges theory and practice.

Reference

Mike Hewitt (2025) Editorial Statement. Transportation Science 60(1):iii-iv. https://doi.org/10.1287/trsc.2025.ed.v60.n1