Crew Rostering Optimization in Long-Distance Freight Railways: A VRP-Based Heuristic for Smarter Railway Crew Management

on

Efficient crew management represents one of the most significant operational challenges in long-distance freight railway systems. Railway operators must simultaneously satisfy labor regulations, maintain continuous train operations, minimize overtime, and efficiently allocate personnel across geographically dispersed depots. Traditional manual rostering methods often struggle to balance these competing objectives, particularly within large-scale freight networks where operational complexity increases rapidly. This study introduces a novel mathematical formulation for the Railway Crew Rostering Problem (CRP) by adapting concepts from the Vehicle Routing Problem with Multiple Depots and Multiple Trips (VRP-MD-MT). Using operational data from Brazil's Vitória–Minas Railway (EFVM), the researchers developed a customized heuristic capable of generating optimized crew rosters that reduce overtime and improve workforce utilization while complying with legal and operational constraints. The research contributes an integrated optimization framework that bridges vehicle routing and railway crew management, providing practical insights for both railway operations research and transportation engineering.

Bibliographic Information

Item Information
Article Title Crew Rostering in Long-Distance Freight Railways: A Multi-Depot VRP-Based Heuristic Approach
Authors Franco Collodetti Mazioli; Rodrigo de Alvarenga Rosa; João Henrique Brunow Barbosa; Hendrigo Venes
Journal Engineering Reports
Volume & Issue Volume 7, Issue 12
Publication Year 2025
Article Number e70511
DOI https://doi.org/10.1002/eng2.70511
Publisher John Wiley & Sons Ltd.
License Creative Commons Attribution License (CC BY)
Online ISSN 2577-8196
Keywords crew rostering problem; railway; railway crew management problem; train driver; vehicle routing problem

1. Research Background

  • Crew rostering is a critical operational challenge in freight railways. Long-distance freight operations require carefully coordinated train driver assignments because individual drivers cannot complete entire journeys within a single work shift. Crew changes must therefore be organized across multiple depots while maintaining uninterrupted railway operations and complying with labor regulations.
  • Labor regulations significantly increase planning complexity. Railway companies must consider maximum shift durations, mandatory rest periods, overtime limits, depot assignments, and rotating work schedules simultaneously. These legal and operational constraints make manual planning increasingly difficult as railway networks expand.
  • Current crew management approaches remain fragmented. Existing research has predominantly focused on the Crew Scheduling Problem (CSP), which determines sequences of tasks, while comparatively fewer studies address the Crew Rostering Problem (CRP), where those tasks are assigned to individual employees across multiple planning periods.
  • Most previous optimization models target passenger railways. Many published methods were developed for passenger railway systems characterized by fixed timetables and predictable service patterns. Freight railways operate under substantially different conditions, including variable demand, longer routes, decentralized depots, and greater operational uncertainty.
  • Integrated optimization remains an important research gap. Previous studies generally optimize only selected aspects of crew management, such as scheduling, depot assignment, fairness, or overtime, without integrating these decisions into a unified optimization framework capable of handling all operational constraints simultaneously.
  • Vehicle Routing Problem (VRP) concepts offer new opportunities. Multi-depot and multi-trip vehicle routing formulations naturally represent depot allocation, route construction, timing restrictions, and resource utilization. Despite these similarities, VRP methodologies have rarely been adapted to railway crew rostering, particularly within freight transportation.
  • The Vitória–Minas Railway provides a realistic industrial case study. The EFVM freight railway in Brazil operates extensive iron ore transportation services involving thousands of train trips each year and approximately 115,000 annual train driver assignments. At the time of the study, roster planning remained largely manual despite its considerable operational complexity.
  • The study addresses both operational efficiency and workforce sustainability. Rather than reducing employment, the proposed optimization strategy seeks to improve workforce allocation by minimizing unnecessary overtime and optimizing driver assignments while maintaining compliance with existing labor legislation.
  • The research bridges transportation optimization and railway operations management. By integrating concepts from the Vehicle Routing Problem with Multiple Depots and Multiple Trips (VRP-MD-MT) into crew rostering, the study establishes a unified mathematical framework that simultaneously models depot allocation, shift construction, legal constraints, and operational cost minimization.

2. Research Objective

  • To develop a unified mathematical formulation for the Railway Crew Rostering Problem by adapting the Vehicle Routing Problem with Multiple Depots and Multiple Trips (VRP-MD-MT).
  • To design a customized heuristic algorithm capable of efficiently solving large-scale crew rostering problems encountered in long-distance freight railway operations where exact optimization methods become computationally impractical.
  • To integrate crew depot allocation, shift construction, overtime management, legal working-time restrictions, and workforce assignment into a single optimization framework.
  • To validate the proposed optimization approach using real operational data obtained from the Vitória–Minas Railway (EFVM), one of Brazil's largest freight railway systems.
  • To compare the optimized crew rostering solutions with the company's existing manual planning approach in terms of workforce utilization, overtime reduction, and operational efficiency.
  • To evaluate the management implications of optimized crew rostering, including potential reductions in operating costs while preserving the existing workforce.
  • To demonstrate the feasibility of applying vehicle routing concepts to large-scale freight railway crew management as a standardized planning methodology.

3. Why This Research Matters

    4. Research Methodology

    • Research Design

      The study adopted a quantitative operations research approach to develop and evaluate an optimization model for the Railway Crew Rostering Problem (CRP). The researchers formulated the problem mathematically by adapting the Vehicle Routing Problem with Multiple Depots and Multiple Trips (VRP-MD-MT) and complemented the formulation with a customized heuristic algorithm capable of solving large-scale real-world freight railway instances within practical computational time.

    • Case Study

      The proposed methodology was validated using operational data obtained from the Vitória–Minas Railway (EFVM), one of Brazil's largest long-distance freight railway systems. The railway primarily transports iron ore and performs more than 16,500 train trips annually, requiring approximately 115,000 train driver assignments. The relatively stable freight demand provided a realistic environment for evaluating crew rostering performance over weekly and biweekly planning horizons.

    • Problem Formulation

      The Crew Rostering Problem was reformulated by representing train drivers as depots, train trips as customers, and driver shifts as vehicle routes. Each route begins and ends at the driver's assigned depot while satisfying legal working-hour regulations, mandatory rest periods, overtime limitations, and depot transition requirements. This adaptation enabled the integration of crew assignment and routing decisions within a single optimization framework.

    • Mathematical Optimization Model

      A Mixed-Integer Linear Programming (MILP)-based mathematical model was developed to formalize the optimization problem. The objective function minimizes the combined costs associated with train driver salaries and overtime payments while ensuring that every train trip is assigned exactly once. The model incorporates numerous operational constraints governing shift duration, depot allocation, crew availability, legal rest periods, overtime limits, crew movement between depots, and workforce scheduling continuity.

    • Customized Heuristic Algorithm

      Recognizing that exact optimization methods become computationally impractical for large-scale railway networks, the researchers designed a specialized heuristic algorithm to efficiently generate high-quality crew rosters. The heuristic seeks feasible solutions that satisfy all operational and legal constraints while significantly reducing computational complexity compared with solving the complete mathematical model directly.

    • Operational Constraints Considered

      The optimization framework explicitly incorporates the principal operational characteristics of long-distance freight railway operations. These include maximum six-hour working shifts, legally mandated overtime limits, minimum daily rest periods, mandatory day-off intervals after consecutive shifts, crew depot assignments, train driver qualification for designated railway sections, and the requirement that all crew changes occur only at authorized crew depots.

    • Input Data

      The optimization model utilized detailed operational information supplied by the railway company. Input data included train schedules, trip origins and destinations, crew depot locations, train driver assignments, depot staffing levels, average train travel times, inter-depot distances, legal labor regulations, shift patterns, salary structures, and overtime compensation rules.

    • Performance Evaluation

      The proposed optimization approach was evaluated by comparing its solutions with the railway company's existing manual planning process. Performance indicators included the number of train drivers required, total overtime hours, workforce utilization, operational feasibility, computational efficiency, and potential annual cost savings resulting from improved crew allocation.

    • Management Analysis

      Beyond optimization performance, the study examined the broader management implications of the proposed methodology. The researchers analyzed how changes in shift duration, depot allocation, workforce distribution, and overtime policies influence operational costs while maintaining compliance with labor legislation and preserving the existing workforce.

    5. Key Findings

    The VRP-Based Formulation Successfully Models Railway Crew Rostering

    The study demonstrates that the Vehicle Routing Problem with Multiple Depots and Multiple Trips (VRP-MD-MT) provides an effective mathematical foundation for representing the Railway Crew Rostering Problem. By modeling train drivers as depots, train trips as routing tasks, and work shifts as vehicle routes, the proposed formulation integrates several operational decisions that have traditionally been optimized separately.

    The unified framework successfully represents depot allocation, shift sequencing, overtime management, crew movement, and labor regulations within a single optimization model while preserving operational feasibility across large freight railway networks.

    The Customized Heuristic Produces High-Quality Solutions for Large-Scale Operations

    The proposed heuristic algorithm effectively solves crew rostering instances derived from a real industrial railway system. Because exact mathematical optimization becomes computationally expensive as network size increases, the heuristic provides an efficient alternative capable of generating practical crew rosters within acceptable computational time while maintaining solution quality.

    The results demonstrate that heuristic optimization offers a practical planning tool for large freight railways where manual scheduling remains the predominant operational practice.

    Optimized Rosters Reduce the Number of Required Driver Assignments

    Compared with the company's existing manual planning approach, the optimization model consistently reduced the number of train driver assignments required over the evaluated planning horizons. These reductions resulted from more efficient allocation of the existing workforce rather than reductions in employee numbers.

    The findings indicate that improved scheduling efficiency enables railway operators to accomplish the same transportation demand using more balanced driver utilization and fewer unnecessary crew reallocations.

    Overtime Was Significantly Reduced Across Multiple Scenarios

    One of the most important outcomes of the proposed methodology is the substantial reduction in overtime hours. By optimizing shift construction and depot assignments simultaneously, the model minimizes situations requiring extended working hours while maintaining compliance with legal working-time regulations.

    Reduced overtime not only lowers operating costs but also improves workforce management by limiting excessive work schedules that may affect employee wellbeing and operational reliability.

    The Optimization Framework Generates Meaningful Cost Savings

    Management analysis indicates that the optimized crew rosters have the potential to generate significant annual financial savings through reduced overtime payments and more efficient workforce allocation. Importantly, these economic benefits are achieved without reducing the size of the workforce, emphasizing productivity improvements rather than personnel reductions.

    The study therefore demonstrates that optimization-based planning can improve both operational efficiency and financial performance simultaneously.

    The Model Fully Respects Operational and Legal Constraints

    All generated crew rosters satisfy the legal and operational requirements governing railway operations, including shift duration, mandatory rest periods, overtime limitations, depot assignments, and crew-change locations. This demonstrates that optimization can improve efficiency without compromising regulatory compliance or operational safety.

    The incorporation of labor legislation directly into the optimization model makes the proposed framework suitable for practical industrial implementation.

    The Framework Standardizes Crew Planning

    The research shows that replacing manual scheduling with a mathematically based optimization framework provides a standardized decision-support process for railway crew management. Automated optimization improves consistency, transparency, and repeatability while reducing dependence on manual experience and subjective planning decisions.

    This standardization supports more reliable operational planning across large railway systems where scheduling complexity continues to increase.

    6. Scientific Contribution

    • Introduces the first unified adaptation of the Vehicle Routing Problem with Multiple Depots and Multiple Trips (VRP-MD-MT) for the Railway Crew Rostering Problem in long-distance freight railways.
    • Develops a comprehensive mathematical optimization model that simultaneously integrates depot allocation, crew assignment, shift construction, overtime control, labor regulations, and workforce utilization within a single optimization framework.
    • Presents a customized heuristic algorithm capable of efficiently solving large-scale industrial crew rostering problems where exact optimization approaches become computationally infeasible.
    • Demonstrates the practical integration of transportation logistics and railway operations research by successfully adapting vehicle routing concepts to workforce scheduling problems in freight transportation.
    • Provides empirical validation using real industrial railway data, strengthening the practical relevance and applicability of the proposed optimization methodology.
    • Expands the methodological foundation of railway optimization research by addressing an area that has received comparatively limited attention compared with passenger railway scheduling problems.
    • Establishes a scalable optimization framework that can serve as a benchmark for future research involving intelligent railway scheduling, automated workforce planning, and integrated transportation optimization.

    7. Industrial Implications

    • Improves freight railway operational efficiency. Railway companies can optimize train driver allocation while maintaining continuous freight operations across geographically distributed depot networks.
    • Reduces overtime-related operating costs. Better shift planning minimizes unnecessary overtime payments without reducing workforce size, contributing to more sustainable operational expenditure.
    • Supports standardized workforce planning. Automated optimization replaces labor-intensive manual scheduling with a consistent, transparent, and repeatable decision-support process.
    • Enhances compliance with labor regulations. The optimization model explicitly incorporates legal requirements governing working hours, mandatory rest periods, overtime limitations, and crew rotation schedules, reducing the risk of non-compliant workforce assignments.
    • Facilitates digital transformation in railway operations. The proposed optimization framework provides a foundation for integrating advanced scheduling algorithms into modern railway management systems and digital decision-support platforms.
    • Supports sustainable human resource management. More balanced workload distribution improves workforce utilization while preserving employee numbers and reducing excessive overtime demands.
    • Offers transferability to other freight railway systems. Because the mathematical formulation is based on adaptable operational parameters, the methodology has potential applicability to other long-distance freight railways operating under similar legal and logistical constraints worldwide.

    8. Research Limitations

    • The study is validated using a single freight railway case study. The proposed optimization framework was evaluated exclusively using operational data from the Vitória–Minas Railway (EFVM) in Brazil. Although this railway represents a large-scale freight transportation system, additional validation using different railway networks would strengthen the generalizability of the proposed methodology.
    • The optimization model reflects the operational regulations of the study context. The mathematical formulation incorporates labor regulations, shift durations, overtime limits, and crew management policies applicable to the Brazilian railway system. Railway operators in other countries may require modifications to accommodate different legal and operational requirements.
    • The research focuses specifically on the Crew Rostering Problem. Other important railway planning activities, such as locomotive scheduling, train timetabling, network capacity management, maintenance scheduling, and disruption recovery, are beyond the scope of the proposed optimization framework.
    • Demand variability is not the primary focus of the study. The EFVM railway primarily transports iron ore with relatively stable freight demand throughout the year. Consequently, the model was developed for a planning environment where workload fluctuations are limited compared with freight systems experiencing highly dynamic transportation demand.
    • The customized heuristic prioritizes computational efficiency. While the heuristic algorithm efficiently generates high-quality solutions for large-scale instances, heuristic approaches do not guarantee mathematically optimal solutions in every optimization scenario.
    • Human-centered scheduling preferences are not explicitly modeled. The optimization framework primarily emphasizes operational efficiency, legal compliance, and cost reduction. Individual driver preferences, job satisfaction, fatigue perception, and fairness considerations are not included as explicit optimization objectives.
    • Real-time operational disruptions are outside the study scope. Unexpected events such as train delays, equipment failures, adverse weather conditions, or emergency operational changes are not incorporated into the current optimization framework, which focuses on planned crew rostering.

    9. Future Research Opportunities

    • Extend the proposed optimization framework to freight railway systems operating under different labor regulations, operational policies, and network configurations to evaluate its adaptability and robustness.
    • Develop integrated optimization models that simultaneously address crew rostering, locomotive scheduling, train timetabling, and maintenance planning within a unified decision-support framework.
    • Incorporate real-time operational disruption management so that optimized crew rosters can be dynamically adjusted in response to train delays, infrastructure failures, and unforeseen operational events.
    • Investigate hybrid optimization approaches that combine the proposed heuristic with advanced metaheuristic techniques, machine learning, or artificial intelligence to further improve solution quality and computational performance.
    • Expand the optimization objectives by incorporating workforce fairness, employee preferences, fatigue management, and quality-of-life indicators alongside operational efficiency and cost minimization.
    • Evaluate the proposed methodology under more dynamic freight transportation environments characterized by fluctuating demand, seasonal operations, and variable train schedules.
    • Develop decision-support systems capable of integrating the optimization model with digital railway management platforms to support intelligent scheduling and real-time operational planning.
    • Investigate the application of the proposed VRP-based crew rostering framework to other transportation sectors, including passenger railways, urban transit systems, metro operations, and multimodal freight logistics.

    10. Potential for Public Policy Citation

    The findings of this study have relevance for public agencies, railway regulators, and transportation policymakers responsible for improving the operational efficiency and sustainability of freight railway systems. The proposed optimization framework demonstrates that workforce productivity and operational performance can be enhanced through intelligent planning rather than workforce reduction, supporting policies that encourage digital transformation within transportation infrastructure management.

    Because the mathematical model explicitly incorporates legal working-hour regulations, mandatory rest periods, overtime limitations, and depot assignments, the methodology provides evidence supporting labor-compliant operational planning. Policymakers responsible for transportation safety and labor regulation may therefore consider optimization-based decision-support systems as practical instruments for improving regulatory compliance while maintaining efficient railway operations.

    The research also contributes to broader transportation policy objectives concerning sustainable freight mobility, infrastructure productivity, operational resilience, and digitalization. National railway authorities and freight operators could use similar optimization frameworks to support long-term workforce planning, improve service reliability, reduce operational costs, and strengthen evidence-based management of critical transportation infrastructure.

    11. Who Should Read This Paper?

    • Researchers in operations research, transportation engineering, industrial engineering, logistics optimization, and railway systems interested in mathematical modeling and large-scale optimization.
    • Railway operations managers responsible for workforce planning, crew scheduling, depot management, and freight transportation efficiency.
    • Industrial engineers developing optimization models for transportation systems, resource allocation, and operational planning.
    • Transportation planners and infrastructure managers seeking analytical approaches for improving operational productivity while maintaining regulatory compliance.
    • Software developers and decision-support system designers creating intelligent scheduling platforms for railway and logistics operations.
    • Government agencies and railway regulatory authorities involved in transportation policy, labor regulation, railway modernization, and digital transformation initiatives.
    • Graduate students and academic practitioners studying optimization algorithms, vehicle routing, railway operations, freight transportation, and workforce management.

    12. Final Thoughts

    This research presents an important advancement in railway operations research by introducing a unified optimization framework for the Crew Rostering Problem based on the Vehicle Routing Problem with Multiple Depots and Multiple Trips. The study demonstrates that concepts traditionally associated with logistics and vehicle routing can be successfully adapted to workforce planning in complex freight railway environments, thereby expanding the methodological foundations of transportation optimization.

    The proposed mathematical formulation and customized heuristic algorithm effectively integrate depot allocation, shift construction, legal working-hour restrictions, overtime management, and workforce assignment into a single decision-support framework. Validation using operational data from the Vitória–Minas Railway illustrates that the methodology can reduce driver assignments required for planning periods, substantially decrease overtime, and improve overall operational efficiency while maintaining compliance with labor regulations and preserving the existing workforce.

    Beyond its immediate contribution to railway crew management, the study establishes a scalable optimization approach that offers practical value for transportation industries seeking to modernize workforce planning through data-driven decision-making. Its integration of mathematical optimization with real industrial operations provides a solid foundation for future developments in intelligent railway scheduling, digital transportation management, and sustainable freight logistics.

  • Addresses an important gap in freight railway optimization. The study extends optimization research beyond traditional passenger railway applications by focusing specifically on the unique operational challenges associated with long-distance freight transportation.
  • Introduces an innovative modeling perspective. Adapting the VRP-MD-MT framework to crew rostering creates a novel optimization methodology capable of simultaneously addressing routing, depot allocation, shift scheduling, and labor regulations within a unified mathematical model.
  • Improves operational efficiency. More effective crew allocation reduces unnecessary overtime, minimizes inefficient driver movements between depots, and supports better utilization of existing human resources.
  • Supports legally compliant workforce planning. The proposed optimization framework explicitly incorporates working-hour limitations, mandatory rest periods, overtime restrictions, and depot assignments, enabling roster generation that satisfies operational and regulatory requirements simultaneously.
  • Provides measurable economic benefits. By reducing overtime payments and optimizing workforce deployment without reducing staff numbers, the methodology offers railway operators opportunities to achieve meaningful annual cost savings through improved planning rather than workforce reduction.
  • Enhances decision support for railway management. Automated optimization replaces labor-intensive manual planning with a standardized analytical framework capable of generating consistent, transparent, and scalable crew rosters for large railway networks.
  • Expands the application of operations research in transportation engineering. The integration of vehicle routing concepts into railway crew management demonstrates how optimization techniques from logistics can be successfully adapted to solve complex workforce planning problems in rail transportation.
  • Provides a foundation for future intelligent railway systems. The proposed framework establishes a scalable optimization model that can support future developments in digital railway operations, automated scheduling systems, and advanced decision-support tools for freight transportation management.

13. Suggested Citations

Teknomekanik (UNP) Style

Mazioli FC, Rosa RA, Barbosa JHB, Venes H. Crew rostering in long-distance freight railways: A multi-depot VRP-based heuristic approach. Engineering Reports. 2025;7(12):e70511. https://doi.org/10.1002/eng2.70511

APA (7th Edition)

Mazioli, F. C., Rosa, R. A., Barbosa, J. H. B., & Venes, H. (2025). Crew rostering in long-distance freight railways: A multi-depot VRP-based heuristic approach. Engineering Reports, 7(12), e70511. https://doi.org/10.1002/eng2.70511

IEEE Style

F. C. Mazioli, R. A. Rosa, J. H. B. Barbosa, and H. Venes, "Crew rostering in long-distance freight railways: A multi-depot VRP-based heuristic approach," Engineering Reports, vol. 7, no. 12, Art. no. e70511, 2025, doi: 10.1002/eng2.70511.

Harvard Style

Mazioli, F.C., Rosa, R.A., Barbosa, J.H.B. & Venes, H. (2025) 'Crew rostering in long-distance freight railways: A multi-depot VRP-based heuristic approach', Engineering Reports, 7(12), e70511. Available at: https://doi.org/10.1002/eng2.70511.

Vancouver Style

Mazioli FC, Rosa RA, Barbosa JHB, Venes H. Crew rostering in long-distance freight railways: A multi-depot VRP-based heuristic approach. Engineering Reports. 2025;7(12):e70511. doi:10.1002/eng2.70511.

Chicago (Author–Date)

Mazioli, Franco Collodetti, Rodrigo de Alvarenga Rosa, João Henrique Brunow Barbosa, and Hendrigo Venes. 2025. "Crew Rostering in Long-Distance Freight Railways: A Multi-Depot VRP-Based Heuristic Approach." Engineering Reports 7 (12): e70511. https://doi.org/10.1002/eng2.70511.

MLA (9th Edition)

Mazioli, Franco Collodetti, et al. "Crew Rostering in Long-Distance Freight Railways: A Multi-Depot VRP-Based Heuristic Approach." Engineering Reports, vol. 7, no. 12, 2025, article e70511. Wiley, https://doi.org/10.1002/eng2.70511.

14. Editorial Note

This review is based exclusively on the scientific content presented in the published article Crew Rostering in Long-Distance Freight Railways: A Multi-Depot VRP-Based Heuristic Approach published in Engineering Reports. The bibliographic metadata were verified using the official publisher webpage, while the scientific interpretation, methodology, findings, contributions, limitations, and implications were derived solely from the published article. This review is intended for educational and scholarly communication purposes and does not replace the original publication. Readers are encouraged to consult the original article for complete mathematical formulations, algorithmic details, and experimental results.

15. SEO Meta Description

Discover how a novel VRP-based heuristic optimizes railway crew rostering in long-distance freight railways. This review explains the mathematical model, methodology, operational benefits, industrial implications, and scientific contributions of a real-world case study conducted on Brazil's Vitória–Minas Railway.

16. SEO Keywords

crew rostering problem, railway crew management, freight railway optimization, vehicle routing problem, VRP-MD-MT, railway operations research, train driver scheduling, railway optimization, crew scheduling heuristic, freight transportation, mathematical optimization, mixed integer programming, transportation engineering, railway logistics, railway workforce planning, depot allocation, operational research, Engineering Reports, railway decision support, freight rail operations

Comments

Post a Comment