CSX Solves Railroad Operation Challenges with and without AnyLogic Rail Library

CSX is a railroad company that operates about 21,000 route miles (34,000 km), including one of the three Class I railroads, which serves most of the East Coast of the United States, and reaches nearly two-thirds of the country’s population.

The Network Planning division’s role is critical to the company’s success. This division identifies where to add capacity to accommodate future growth, ensures infrastructure can support and sustain a high level of service, and tries to improve the efficiency of capital spending.

Network Planning uses a multistep approach to manage the network capability. They utilize analytical tools to monitor current service levels, identify appearing problems, and determine the root cause of these disruptions (if the problem is operational or infrastructural).

In addition, they analyze what possible solutions can be applied to the problem, including investment decisions, and which of these decisions will provide the best financial return. To get the right answers, the use of traditional analytical tools is insufficient. That is why, for these purposes, CSX employs simulation modeling technologies. They use AnyLogic software for many different purposes because it allows them to create models of various systems, at the required abstraction level, with a quick turnaround time.

AnyLogic allows the railroad industry users to simulate line-of-road, terminal, and yard problems. The following three projects, completed by CSX in 2014, covered a variety of tasks that were solved using AnyLogic software.

MGA Line Investment Planning


A rail line that is jointly owned by CSX and their competitors was expected to see a large growth in demand from several coal mines. The high competition between the two companies meant that if one of them could not fulfill the demand, the other one would do so. CSX needed to identify the best operational/capital strategy to handle the increased business. They wanted to know the answers to these specific questions:

  • Did they have enough staging capacity on the line to stage empty unit coal trains to respond quickly to the new demand?  
  • Where were the best locations on the line to add the additional staging capacity if needed?

They utilized AnyLogic simulation modeling to find the answers.


Rail Line Simulation Model

MGA Rail Line Simulation

The created supply-chain network model simulated the demand of empty trains from five coal mines, as well as the fulfillment of the demand, and staging of empty trains. The trains were modeled as agents moving across the network. By varying values of relevant parameters, users could infer the impacts of different factors to the train throughput (i.e. staging capacity, as well as loading speeds at the coal mines).

The model calculated the company’s achieved throughput, and the business lost by CSX, due to the lack of available trains.


The model provided a way for decision makers to gain insight into the system to help identify the maximum possible throughput. The simulation showed that the company did not have enough staging capacity to serve the increased demand, and it helped distinguish the highest priority capital investment projects to implement.

Locomotive Shop Optimization

Locomotive Shop Simulation Model

Nashville Locomotive Shop Redesign


The CSX’s Nashville locomotive shop needed to be expanded in order to meet the higher level railroad network redesign. The facility included a quality maintenance shop and a roundhouse. The company’s mechanical department needed to select the best layout design from eight alternatives. The objective was to identify the layout that maximized the throughput of locomotive processing.


This project utilized the special AnyLogic Rail Library to build a model of the locomotive shop and test the different designs.

In the model, 72% of the incoming locomotives went to the roundhouse, while 22% went to the maintenance shop. The remaining 6% could go to either of them, depending on the problem they had after further inspection. Service times in both shops differed.

Locomotives moved at five miles per hour in the system. There was one common queue, with nine spots, for both shops. A locomotive was pulled into the system if there was a spot available in the roundhouse, the maintenance shop, or the common queue. The numbers of spots available in both shops and in the queue were parameters that could be varied by the user.


The model was used by the mechanical department to test their assumptions by experimenting with the system, and as a decision support tool to determine which layout configuration was best. The model helped the specialists drive the conversation among the stakeholders and base their solution on the reliable data.

Network Performance Emulator

The company faced greater than expected demand growth, coupled with hard winter weather and resource constraints, which led to congestion on the northern tier of the CSX network. When they analyzed this problematic situation afterwards, the Network Planning team was trying to determine what happened on the network and could avoid these issues in the future.

As the research continued, they found out that it would be easier to understand the processes if they replaced the traditional analytical methods with a visual emulator. So, they decided to reproduce, or replay, the past system behavior in AnyLogic with the use of animation on a GIS map to better understand density, flow, and congestion processes in the network and improve decision making. All of the train movement data was imported to AnyLogic from the databases, predefining the behavior of the trains in the model. The emulator included the animated train movement with statistics and indicators making the data visually understandable.

The model was presented to the C-level officers and the customers and helped drastically raise the understanding of the issue among the stakeholders.

Rail Network Simulation

Rail Network Emulator

Watch the video of Jeremiah Dirnberger from CSX presenting these case studies at the AnyLogic Conference 2014:

More Case Studies

  • Simulation de la logistique du rail interne du Port du Havre
    Le port du Havre avait besoin d’aide pour construire un nouveau terminal multimodal. Il s’agissait de simuler le transfert des conteneurs entre les trains / péniches fluviales et les autres terminaux existants, qui transféraient eux-mêmes ces conteneurs en direction ou en provenance du transport maritime.
  • Optimisation du réseau ferroviaire français
    Le gestionnaire du réseau ferroviaire national voulait savoir si le transport de fret par rail pouvait être compétitif face au transport par camions. Il voulait rendre le transport par la voie camion-rail-camion plus efficace.
  • Simulation de la construction d’un tunnel à l’aide d’un tunnelier
    Le coût d’une heure d’arrêt d’un tunnelier (machine de forage de tunnel) est habituellement élevé et les gestionnaires de projet doivent s’efforcer d’éviter les retards dans la construction. Le but du projet de simulation, qui a été mené à l’université Bochum de la Ruhr en Allemagne, consistait à créer un modèle de simulation capable de déterminer les goulots d’étranglement dans les processus de construction du tunnel, afin de minimiser les pertes financières potentielles.
  • Modélisation de la capacité d’une gare de triage
    Aurizon est le premier opérateur de fret ferroviaire d’Australie, à la tête de plus de 700 locomotives et 16 000 wagons. Aurizon prend en charge le transport du charbon, du minerai de fer et des minéraux à grand échelle. Afin d’augmenter l’efficacité opérationnelle, l’entreprise a décidé de déplacer l’une de ses gares de triage dans une autre ville. Cette gare de triage gérait essentiellement l’entretien des wagons et des locomotives, ainsi que la préparation des locomotives.
  • Effet “coup de fouet” dans la chaîne d’approvisionnement des semi-conducteurs
    La direction de la chaîne d’approvisionnement d’Infineon, un gros fabricant de semi-conducteurs, voulait étudier l’effet coup de fouet affectant son marché afin de diminuer les dépenses et de mieux prévoir le comportement du marché. A l’aide du logiciel AnyLogic, elle a construit un modèle de chaîne d’approvisionnement, depuis les matières premières jusqu’au marché.
  • Sélection de la meilleure politique de gestion des stocks à l’aide de Gojii
    Les outils existants de gestion de la chaîne d’approvisionnement et les S&OP (Ventes et planning opérationnel) sont performants lorsqu’il s’agit de gérer l’approvisionnement afin de répondre à une «prévision» sélectionnée. Cependant, il n’y a pas de «prévision correcte» de la demande future, et les outils existants ne sont pas conçus pour sélectionner le meilleur niveau de demande pour l’entreprise. Il existe une «insuffisance d’outil» entre les données entrées sur les prévisions et la sélection du meilleur signal de demande (la "Prévision") pour gérer votre système S&OP. Gojii est l’outil créé par DecisioTech pour combler cette lacune.
  • Planification du réseau de distribution et optimisation du stock à l’aide de simulations
    Diageo est une entreprise multinationale britannique de boissons alcoolisées. Diageo Russie est l’un des cinq distributeurs majeurs de Russie de boissons alcoolisées en gros, une activité qui génère traditionnellement peu de marges, et dans laquelle le bénéfice est tributaire du niveau de service client et de coûts logistiques élevés. La société Diageo a cherché assistance auprès d’une société de conseil Amalgama, LLC, lorsqu’elle a dû faire face à une augmentation de son volume de ventes, sans pour autant réaliser de meilleurs profits du fait des coûts logistiques par unité.
  • Modélisation d’un réseau de transport centré sur le client
    La sphère des services de transports publics en Australie fait l’objet d’une transformation, afin de répondre à des changements démographiques. Afin de mieux répondre à ces nouveaux défis, les compagnies de transport publiques doivent comprendre le comportement de leurs réseaux du point de vue du consommateur.
  • Design and Analysis of Marine System for Oil Transportation in the Arctic by Means of Simulation
    The Novoportovskoye oil and gas condensate field is located in the Yamal peninsula and is owned by Gazprom Neft, the fourth largest oil company in Russia. Oil from the field is transferred via a 100km pipeline to the sea terminal at Cape Kamenny (the Gulf of the Ob river), where it is loaded into arctic shuttle tankers for further transportation to Murmansk. The full field development will start in 2016 and continue for several decades.
  • Preventing “Bus Bunching” with Smart Phone Application Implementation
    In public transport, bus bunching refers to a group of two or more transit vehicles (such as buses or trains), which were scheduled to be evenly spaced running along the same route, instead running in the same location at the same time. Dave Sprogis, Volunteer Software Developer, and Data Analyst in Watertown, MA, used AnyLogic to confirm his thesis that preventing "Bus Bunching" would improve the experience of public transit bus riders.