OV SAAL Zuidtak Oost is a large construction site where BAM Rail expands rail capacity from two to four railway tracks over 8 kilometre. In order to have a safe work environment, BAM Rail is allowed to possess the railway tracks for variable durations. For OV SAAL Zuidtak Oost a 9 day-rail possession is scheduled for the beginning of August 2016. For these nine days, at four different connected construction sites rail related construction works are executed. A strict and complex planning schedule is needed to guarantee all construction works can be completed before the deadline. In preparation for rail possessions, Gantt-charts and time-distance diagrams are used to create the planning schedules. Gantt-charts consist of a chain of linked activities representing the different tasks the construction site managers have to execute. Problem statement The planning methods that are currently used provide insufficient overview over complex TVPs (‘TreinVrije Perioden’). Only very experienced planners have the knowledge to make and interpret these planning schedules. Users identify three missing aspects resulting in failure costs and inefficiencies: The currently used planning methods show no spatial overview of the state of the complete line system at each moment of time. No clashes between different operational activities can be detected. Construction logistics (working trains, heavy machinery) cannot be represented with the current methods. BAM Rail experiences failure costs and expensive operational inefficiencies by the application of Gantt-charts and time-distance diagrams for general rail projects. Many users indicate that a spatial representation of a Gantt-chart would be useful in eliminating risks and inefficiencies. Therefore, the objective of this research is described as follows: To design a planning visualization method which assists BAM Rail in enhancing insight into the state of a linear construction site over time. Apart from the design of a planning visualization method, this research presents several additional recommendations related to observed processes during rail related construction works. Current situation User interviews and site visits were conducted to perform an in-depth analysis of the current situation. After categorizing the users’ responses, six main causes of risks and inefficiencies were identified: 1. Absence of clash detection: some clashing activities and invalid state sequences are left unidentified in the planning phase, resulting in failure costs during the construction phase. 2. High project complexity: Gantt-charts become too complex to comprehend, resulting in experts making mistakes when creating planning schedules. 3. Missing project overview: no basic site overview is available, resulting in non-expert decision makers making wrong decisions. 4. The impossibility to perform a scenario analysis: this results in failure costs as no control measures and fall back scenarios can be quantitatively analysed. 5. The impossibility to show construction logistics real-time: train paths and machinery locations are unknown, resulting in failure costs when making operational decisions. 6. The impossibility to determine delay propagations real-time: measures to settle delayed activities cannot be quantitatively analysed real-time, resulting in failure costs when making operational decisions. A requirements analysis presents a set of requirements for a visualization method. The complete set of requirements spans a solution space for a high-level design. High-level design Stakeholder interviews and test runs resulted in insights in the desired model output. The desired model output has been used to create a platform-independent high-level design. The high-level design consists of a model architecture, software architecture, pseudo code and system specification for use in the implementation in a software package. An important aspect of the high-level design is clash detection. Five different types of clashes are identified. For all five an automatic clash detection module is required. Activities moving in time are included in the high-level design as well. Implementation The suitability for implementation of the high-level design has been evaluated for two software packages, Navisworks and ArcGIS. The four main categories of evaluation criteria are defined as functional criteria, project management criteria, user interface criteria and configuration criteria. Preliminary tests with both software packages formed the basis for the evaluation. ArcGIS outperformed Navisworks in three out of four main criteria. After a sensitivity analysis, it was decided to create a prototype of the high-level design in ArcGIS. During the process of generating and validating the desired user output in ArcGIS, several shortcomings of ArcGIS were detected. These shortcomings were sorted based on the main functional aspects of the visualization model. Software developer Esri Nederland BV has been consulted to confirm the identified shortcomings and asked for strategies to overcome these limitations. Several recommended adjustments for ArcGIS are presented to Esri before a fully functional model can be implemented. Esri estimated that an investment of --------- and -------- are necessary to deliver a fully functional visualization model, satisfying all user requirements. Research conclusions The conclusions of this research answer the main research question of this research: How can BAM Rail get insight into the state of the construction works at a line construction site at each time step? Firstly, it can be concluded that failure costs are associated with the identified risks and inefficiencies. Providing an adequate insight into the state of the construction site system at each time step reduces the planning related risks and thus the failure costs. It can be concluded that the state of the construction works can be represented best by several aspects: 1. For each planned activity the location must be presented in a geospatial overview. Both static and dynamic activities must be included. 2. For each user-defined track section the state of that section must be presented in the geospatial overview. 3. Time and infrastructure constraints must be included in the geospatial overview. It is concluded that a planning visualization method is necessary to get insights into the state of the construction works of a line system for each time step. A module to automatically detect clashes is necessary to assist users in detecting the identified types of clashes. The high-level design included how the model architecture and software architecture of a planning visualization model should look like. By implementing a software package with these characteristics and structure, insights into the state of a construction site system can be provided. Esri’s ArcGIS is found to be most suitable for the implementation. The possibility to present activities and states as polygons matches with the desired user output. Furthermore, CAD drawings and aerial photos facilitate an overview on scale. Finally, intuitive navigation through ArcGIS is considered as user-friendly. Concluding, creating a planning visualization model in ArcGIS results in satisfying the user requirements– from planner to site manager – and thus solve the six main causes of risks and inefficiencies. Recommendations BAM Rail is advised to request an official proposal from Esri for a visualization method in ArcGIS. In addition, a BAM project team leading the implementation in ArcGIS should be formed. This team is responsible for monitoring the progress and deadlines, organizing test runs and defining milestones and showstoppers. Additional recommendations are presented. First, two recommendations for BAM Rail are presented. Then, recommendations for further research are presented. It is observed that the communication between the coordinating construction site manager (BCU) and construction site managers (‘uitvoerder’) during the construction works is done only by phone, once every 2-4 hours. Users consider these phone calls as inefficient for regular operations. Therefore it is recommended to investigate the implementation of an integrated mobile system for site managers to report the progress using an application on a mobile phone or tablet. Furthermore, it is recommended that a more thorough workplace inspection is executed before a TVP starts. Currently, it often occurs that physical obstacles or deviating situations are only discovered during the construction phase, resulting in a wrong initial state. When these deviations are discovered in an early phase, failure costs can be prevented. This research presents a visualization algorithm rather than an optimization algorithm. It is recommended to investigate the possibility of implementing an optimization algorithm to not only visualize planners’ decisions, but also making them. Two aspects must be changed when designing an optimization algorithm. First, the addition of a probabilistic activity duration instead of a deterministic activity duration must be added. Second, expressing the starting time of each activity in completion time of prior activities must be added. Finally, it is recommended to research the possibility of implementing resources in the model. By considering available working crew, machinery or costs, the use of these materials can be viewed at each moment in time. Also for the extra resources constraints can be defined. Resources can be implemented in the clash detection model as well.