The Railway Bridge Game usability, usefulness, and potential usage for railways management

Transcription

1 The Railway Bridge Game usability, usefulness, and potential usage for railways management Rens Kortmann Delft University of Technology, The Netherlands Emdzad Sehic ProRail, The Netherlands Abstract. The Railway Bridge Game (RBG) offered Dutch railway infrastructure manager ProRail insight into the value of gaming for its organisation. This paper describes the design and evaluation of the RBG, commissioned by ProRail and the Next Generation Infrastructures (NGI) foundation, in the context of a practical case study of railway service extension. The game was used in a series of game sessions with ProRail employees at the Amsterdam rail traffic control centre. The players behaviour was logged and their attitude towards the RBG was measured through interviews and pre- and post-game questionnaires. The players decided that the usability of the RBG was good. Moreover, they and other ProRail staff members perceived gaming as a useful method and foresaw the usage of the (improved) RBG to 1) improve decision taking in the organisation; 2) support the design of new timetables, physical infrastructure, and collaboration processes; and 3) recruitment and professional training of employees. We concluded that these three ways of using games may benefit railway management organisations and related actors to formulate answers to strategic challenges such as improving service levels, security, and cost efficiency. 1. INTRODUCTION ProRail manages the railways infrastructure in the Netherlands and faces an increasing demand of its services in the coming decades. One of the challenges lies in raising the capacity and service level of the Schiphol Amsterdam Almere Lelystad (SAAL) corridor (MinTrans, 2008). One of the bottlenecks on this corridor that currently obstructs further growth of its capacity is the railway bridge over the Vecht river. Also security and cost efficiency issues play an important role in the search of improved service levels. In this paper, we use the practical case study of the SAAL corridor and the Vecht bridge as a starting point to develop the Railway Bridge Game (RBG). The practical aim of this game was to familiarise ProRail with the theory and practice of gaming to support decision taking, design processes, and professional training. Our research purpose was to estimate how ProRail staff perceived the usefulness of gaming in their organisation. Also, we aimed to determine the potential usage of gaming by ProRail and related actors to formulate answers to strategic challenges such as improving service levels, security, and cost efficiency. To evaluate the RBG, we organised a series of game sessions with a number of relevant participants from ProRail. We recorded the behaviour of the players and provided them with feedback on their performance. Using questionnaires and interviews, we were able to assess the usability and perceived usefulness of the RBG. We used the Technology Acceptance Model (Davis, 1989) to arrive at the attitude towards gaming. Thereafter, we estimated the potential usage of gaming for ProRail and other rail-based organisations through interviews with game players and other ProRail staff. The outline of the paper is as follows. In section 2 we provide the background of the problem and describe the relevant related work. Thereafter, section 3 presents the design of the game. In section 4, we describe the design of a series of game sessions with ProRail employees and the validation of the game. Thereafter, in section 5, we reflected on the usability and usefulness of the RBG and on the potential usage of gaming for ProRail. Finally, we draw conclusions on the attitude towards gaming and the potential usage of gaming for ProRail and related actors in section BACKGROUND AND RELATED WORK The work in this paper is part of a larger research programme in which ProRail, TU Delft, and the Next Generation Infrastructures foundation cooperate. Altogether, three projects were executed in this programme to explore the value of gaming for ProRail. The first project aimed to explore and validate alternative modes of organization for rail cargo capacity management (Meijer et al, 2009a). This project involved the design and evaluation of a management game. The second project studied a control concept for high-frequency train transport (Meijer et al, 2009b). This process involved a multi-player process simulation using existing ICT tools. The third project, on which this paper reports, involved a single-player process simulation using newly developed ICT tools. Together, the three projects provided a clear picture of different types of games and the value of gaming for ProRail. As a background to the third project, we already mentioned the OV-SAAL programme of the Dutch

2 Ministry of Transport and ProRail in section 1. Within this programme, ProRail studied several solutions to relieve the bottleneck caused by the bridge over the Vecht river (Van Es, Murtas, and Snijders, 2008). The authors concluded that the bottleneck is best solved by switching from planned to created bridge openings. Currently, the bridge openings of the Vecht bridge are planned. This means that the timetable is constructed such that there are no trains scheduled on the bridge for a period of 6 minutes every hour. This time window allows the bridge be opened for shipping traffic. In contrast, when using created bridge openings, there is no time window planned to open the bridge. Instead, rail traffic controllers and bridgemen share the responsibility to open the bridge when necessary while minimising the delay for all (shipping and rail) traffic. Switching from planned to created openings enables planners to increase the train frequency over the bridge, since the 6-minutes time window, previously allocated to bridge openings, can now be used for rail traffic. This change, together with other (technical) measures, would allow the frequency of trains passing the bridge to increase from approximately 15 to 23 per hour in each direction. However, the authors recommended that more research would be conducted around the effects of created bridge openings before the switch were implemented in practice. 3. GAME DESIGN To familiarise ProRail with the theory and practice of gaming we designed a game around the problem of the Vecht bridge bottleneck in the OV-Saal corridor. The aim of the game was to explore the effects of switching from planned to created bridge openings, while increasing the frequency of trains passing the bridge. We especially focussed on the effects on the quality of the realised train service and on the work load of traffic controllers. Although the game was not intended to solve it, we nevertheless used this real-life problem to maximally enable the players and other ProRail staff to comprehend the usefulness of the game. Below we described the functional design and implementation of the RBG game. 3.1 Functional design The target players of the RBG are professional train traffic controllers who in real life use a computer system to control rail traffic. The RBG therefore enables single-player, computer-based game play (which is currently being extended to multi-player game play). The functional design is shown in Figure 1. It consists of a player application and a game facilitator application that both contain a number of components. The two applications communicate through a local network. Below we described the components separately. Player application VPT-PRL model Train simulator ARI: automatic allocation of train routes component Communication through local network Game facilitator application Facilitation controls Bridgeman controls Figure 1: functional design of the RBG software VPT-PRL model A rail traffic controller s task is to set train routes for safe and sole use by approaching trains. In real life, traffic controllers use the software system VPT-PRL to perform this task. The game software therefore contains a VPT-PRL model that mimics the real-life system as closely as is needed to successfully play the game scenarios. Our model provides the player with a user interface that consists of four computer screens (see Figure 2). The first screen displays the train schedule (A: plan screen). The second screen shows the controls of individual infrastructure elements and a detailed picture of the study area (B: operate screen). The final two screens display the tracks in the study area, the allocated train routes, and the estimated position of actual trains (C&D: track display screens) Train simulator In order to provide the players with feedback on their actions, a discrete-event train simulator was designed to keep track of the positions of trains in the study area. The simulator generates delays in train movements that resemble the statistics of delays in the study area. The flowchart that describes the simulator behaviour is shown in Figure ARI: automatic allocation of train routes In real life, an automatic route allocation system called ARI supports traffic controllers when the train service is undisturbed. Only when disturbances appear that lead to large delays, traffic controllers need to deactivate ARI and switch to manual allocation of routes. Therefore, we designed a game component to automatically allocate train routes in the same way as the ARI system. The flowcharts describing the behaviour of this component are shown in Figure 4 and Figure 5.

3 A B C D Figure 2: design of the user interface of the player application in the RBG software: plan screen (A); operate screen (B); track display screens (C&D) Figure 3: flowchart of the train simulator used to estimate the positions of trains during the game Figure 4: flowchart of the automatic route allocation component (part 1)

4 Figure 5: flowchart of the automatic route allocation component (part 2) Facilitation controls The system contains controls for the game facilitator to start, pause, resume, and stop the game; to adjust the game speed; and to load different timetables in the game before starting it Bridgeman controls The game facilitator plays the role of bridgeman. Therefore the game facilitator application also provides controls to open and close the bridge. The protocol for opening a bridge is provided in the sequence diagram shown in Figure 6. bridgeman to open the bridge. When the shipping traffic has passed, the bridgeman closes the bridge and notifies the traffic controller, after which the bridge is locked. Optionally, the traffic controller may ask the bridgeman if an opening is required when the train traffic is quiet. 3.2 Implementation We used the hardware configuration shown in Figure 7 to implement the design of the RBG. Figure 6: protocol to handle a bridge opening In the figure, it is shown that a bridgeman requests a bridge opening to the traffic controller (which is usually done by telephone). When a suitable moment arrives, the traffic controller unlocks the bridge to allow the Figure 7: hardware configuration to implement the RBG design As is shown in the figure, four computer screens were connected to a game workstation (low-cost, state-ofthe-art machine with two graphics cards). On the workstation we ran the train simulator, the ARI system, and the game player application. The game facilitator application ran on a separate laptop (game facilitator console) that was connected to the game workstation through a local network. We used the Java 6 Standard Edition programming language to implement the software. The communication between the game facilitator and game player applications was accomplished through Remote Method Invocation (RMI) which is a standard feature of Java.

5 The costs of the entire setup are modest (below 5000) and the hardware can be easily transported in a passenger car or small van. 4. GAME SESSIONS AND VALIDATION In November 2009 we hosted four days of game sessions at the ProRail traffic control centre in Amsterdam. Every day featured a new game scenario. During the game play we measured the player s performance in terms of the acuity of the realised train service and we estimated his/her work load. These measurements and the response to pre- and post-game questionnaires were used to validate the game and to estimate the usability, usefulness and potential usage of the game. 4.1 Setup of the game space Figure 8A shows a real ProRail traffic control system, after which we set up our game space shown in figure B. The game space consists of a desk and four computer screens that present the information depicted in Figure 2. The players interacted with the system using a keyboard and mouse. The facilitator was positioned directly left of the player. A video camera was installed behind the player to record his/her actions for the purpose of analysis. 4.2 Validation We validated the game through questionnaires with experts and by comparing the acuity of the realised train service in the game to the statistical acuity in practice. After the first day of game play we asked our players about the quality of the game and how well they recognised the game interface as a model of the real traffic control systems. The questions were repeated after the fourth day of game play. The quality of the game scored 4.3 ± 0.7 on a 5-point scale after the first day and 4.0 ± 0.7 on the fourth day. On a 10-point scale, resemblance scored 7 ± 2 after the first session and 7 ± 1 after the fourth session. When comparing the acuity of the realised train service in the game to the statistical acuity in practice we obtained the results shown in Figure 9. Delay (minutes) senior player qualified player measured Figure 9: the average delay and standard deviation of the realised timetable of two players (both experienced traffic controllers) and the measured delay from a regular day in the area that was modelled As can be seen, the average delay of two experienced players falls well within the spread of the actual measured delay. We therefore concluded that the realised service in the game is a valid representation of the service realised in practice. A Altogether, the outcomes shown above indicate that the game is a valid representation of rail traffic control in practice. 5. USABILITY, USEFULNESS, AND POTENTIAL USAGE After the game sessions, we interviewed the players about the usability, usefulness, and potential usage of the game. Also, we asked other ProRail staff about the usefulness and potential usage after explaining the game to them and showing them the results and a video presentation of the sessions. Below we summarise their responses. B Figure 8: a real traffic control system used by ProRail (A) and the RBG game space used in our sessions (B) 5.1 Usability The usability of the game reflects how players perceive the ease of using the game (Davis, 1989). According to the players, the RBG shows much resemblance to the VPT-PRL traffic control system that the game was

6 modelled after (score 3.9 on a 5-point scale). Moreover, they judged that the explanation of the game was clear and their tasks were well defined (scores 4.8 and 4.4, respectively). Although the players indicated several possibilities for improvements, we concluded that the usability of the RBG was high. 5.2 Usefulness We draw upon the term usefulness to indicate to what extent the game would enhance the performance of ProRail and its employees. When asked about the RBG, both players and other ProRail staff reported that they perceived the (improved) RBG as a useful tool to improve their personal performance and the performance of ProRail as a whole. Players felt that the game may be especially helpful to learn how to tackle unexpected circumstances, for instance, in the case of incidents on the tracks. They felt motivated by the game and enjoyed playing the RBG. One player even came in on a day off not to miss a game session. Other, more senior staff perceived the game as a valuable addition to existing simulation studies. They reported that human factors play a major role in the quality of a train service. For instance, a well-designed timetable may yield a poor service when the interplay between traffic controllers, train drivers and traveller information providers is not well coordinated. It is very difficult to simulate the richness of human behaviour in computer simulators and the current ProRail simulators are not equipped for this. In contrast, gaming allows human players to participate in computer simulations thus adding to the richness of the simulation (Mayer, 2009). Therefore, gaming offers the opportunity to enhance the validity of simulation studies by placing a human in the loop. 5.3 Potential usage Game players and other ProRail staff foresaw potential usage of the RBG and other types of gaming in roughly three areas of application: 1. Professional training 2. Improvement of decision taking 3. The support of design processes First, the (improved) RBG might be used to recruit and train (new) traffic controllers. Many organisations that struggle to recruit well-qualified personnel use games to interest candidates. Examples are the US Army (Davis, 2004) and Air Traffic Control in the Netherlands (LVNL, 2010) who launched games to attract new employees. Future versions of the RBG might also serve as a tool to test the entry level of applications, which is likely to improve the cost efficiency of the hiring process considerably. Currently, apprentice controllers at ProRail are trained on costly, immobile simulator systems that cannot be reconfigured to simulate other areas than the area that it was originally designed for. In contrast, the RBG is mobile, cost efficient, and will simulate different areas without extensive adaptations of the software. Also more experienced traffic controllers might use the RBG as a training tool, for instance, to review their actions during past incidents, to improve their responses to future incidents, and to prepare for changes in timetables and/or physical infrastructures due to construction works. Second, ProRail analysts foresaw the use of gaming to improve decision taking by the ProRail management. The current simulators typically provide their audience with numbers and some crude visualisations of, for instance, train movements. Conversely, games provide decision takers with lively experiences of future scenarios and the possible outcomes of their decisions. Especially decision takers without a technical background will be better able to comprehend and appreciate the outcomes of gaming studies when compared to pure computer simulations. Therefore, our interviewees expected gaming potentially to be used for improved decision taking in ProRail. Third, the RBG was perceived as a potential tool to support design processes in ProRail: games may serve as ex-ante evaluation tools for new designs. Infrastructure managers continually face strategic challenges such as the improvement of service levels, security, and cost efficiency. Examples are the design of new time tables and new traffic control protocols (this paper; Meijer et al., 2009b), of new physical infrastructure and security systems, and of new information systems for traffic controllers, train drivers and others. Finally, ProRail plans the use of games to (re)design collaboration processes between traffic controllers, train drivers, rolling stock and personnel (re)schedulers, traveller s information providers, emergency services, and other relevant actors. 6. CONCLUSIONS We designed and evaluated the Railway Bridge Game to familiarise Dutch railway infrastructure manager ProRail with the theory and practice of gaming. From the game sessions that were organised at the ProRail traffic control centre in Amsterdam and from the interviews with players and other ProRail staff afterwards, we concluded that ProRail holds a positive attitude towards gaming. Playing the game showed its potential to help solve the bottleneck of the Vecht bridge on the OV-Saal rail corridor. ProRail foresees potential usage of the (improved) RBG and other games in roughly three areas of application: training (new) employees, improving decision taking by the management, and supporting the design of timetables, traffic control protocols, physical infrastructures, and collaboration processes between ProRail, railway companies, emergency services, and other relevant actors. We are currently continuing our collaboration and are working towards an improved version of the RBG that interfaces with the existing train simulators of ProRail. Future research will show how this game may further support ProRail and related actors to formulate

7 answers to strategic challenges such as improving service levels, security, and cost efficiency. REFERENCES Davis, F. (1989) Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), pp Davis, M. (2004) America s Army PC Game Vision and Realization. San Francisco: US Army and Moves Institute, January LVNL (2010) website (in Dutch) last viewed 31 st March MinTrans (2008) Planstudie openbaar vervoer Schiphol Amsterdam Almere Lelystad (OV Saal) (in Dutch). Dutch Ministry of Transport, Public Works, and Water Management. March Mayer, I. (2009) The Gaming of Policy and the Politics of Gaming: A Review. Simulation & Gaming, Dec 2009; vol. 40: pp Meijer, S., Mayer, I., Van Luipen, J., and Weitenberg, N. (2009a) Gaming Rail Cargo Capacity Management: Exploring and Validating Alternative Modes of Organization. In Yeo, G. and Cai, Y. (Eds.) Proceedings of the 40th annual conference of the International Simulation and Gaming Association (ISAGA), Singapore, July 2009, ISBN Meijer, S., Van der Kracht, P., Van Luipen, J., and Schaafsma, A. (2009b) Studying a control concept for high-frequency train transport. Proceedings of the Next Generation Infrastructures Conference, Chennai, India, 9-11 December Van Es, A., Murtas, C., and Snijders, R. (2008) Knelpunt Vechtbrug Weesp (in Dutch). ProRail, December 2008.