There are currently railway stations in the Netherlands [1] including four which are used only during special events and one which serves the National Railway Museum only. NS Stations is the body which manages and owns all railway stations in the Netherlands. Stations are divided into two categories based upon the service they receive. These are, in order of decreasing importance:. There are exceptions to this categorization. Some local trains — despite being called stoptreinen — do not stop at all stations: two examples are the services from Groningen to Roodeschool and from Tiel to Arnhem.

Author:Faukora Kagagor
Language:English (Spanish)
Published (Last):24 June 2012
PDF File Size:20.13 Mb
ePub File Size:12.18 Mb
Price:Free* [*Free Regsitration Required]

Twenty-two train traffic controllers enacted two scenarios in a human-in-the-loop simulator. Their experience, goals, strategic mental models, and performance were assessed through questionnaires and simulator logs. Goals were operationalized through performance indicators and strategic mental models through train completion strategies.

A variation was found between operators for both self-reported primary performance indicators and completion strategies. The level of experience tends to affect performance differently. There is a gap between primary organizational goals and preferred individual goals. Further, the relative strong diversity in primary operator goals and strategic mental models indicates weak resilience at the individual level. With recent and upcoming large-scale changes throughout the sociotechnical space of the railway infrastructure organization, the findings are useful to facilitate future railway traffic control and the development of a resilient system.

For the Dutch railway infrastructure managing organization ProRail, the notion of resilience and robustness strongly resonates in the organization to improve the system along these concepts Meijer, Research often focuses on a specific unit of analysis, as it is not yet well understood how resilience is linked across these different levels Righi et al.

This phenomenon can be also be labeled as a gap between the system as designed or imagined and the system as it is actually operated, which results in a distance between the various levels Dekker, As such, performance variability is normal, though it needs to be controlled.

Performance variability that leads to positive outcomes should be promoted Hollnagel, , Departing from resilience studies in the Dutch railways at a system and organizational level, this study focused on the individual level of railway traffic operators in order to provide recent and quantitative insights to further the understanding of variations in their cognition and behavior and the implications thereof.

The central research questions were as follows: To what extent do organizational and individual goals correspond? The following section briefly introduces the Dutch railway system from a number of perspectives. This overview is followed by a brief presentation of the theoretical background to goals and strategic mental models. The subsequent sections present the method, results, and discussion and conclusion.

For instance, reliability can be conceptualized in a number of ways, such as punctuality, which can be further operationalized in terms of, for example, arrival, departure, or overall arrival and departure punctuality. Departure punctuality was a performance indicator until , when arrival punctuality became the indicator Veeneman, The formalization of performance indicators is an annual iterative process with occasionally ad hoc organizational reactions throughout the year in the case of unexpected large-scale disruptions that are subject to media scrutiny.

Railway traffic operations differ between European countries in a number of ways, such as organization, roles and responsibilities, and level of automation Golightly et al. In the Netherlands, a train traffic management system is used to execute the timetables, which are operated by train traffic controllers.

The primary responsibility of these controllers is to execute train timetables in an accurate and punctual manner Sulmann, However, a more active role is needed in unsafe situations that cannot be controlled by the automated safety system or when there is a system malfunction Sulmann, Given the restriction of a capacity increase through the mere addition of tracks, a change in the organizational processes is also required. As such, process optimization programs are being implemented that focus on, for instance, increasing the centralization of decision making to the national control center operational control center rail [OCCR] for disruption mitigation procedures and restructuring the roles and responsibilities of operators.

Switches are increasingly being removed at major stations e. Finally, the replacement of the current traffic management system is being explored. In a dynamic environment, individuals focus on elements in the environment that are goal related. Deriving the meaning of the elements and the projection to the future is done in light of the goal and the active mental models Endsley, Goals influence the valuation of multiple options during decision making Mohammed et al.

In order to achieve resilience, operators need to have a common set of goals Lengnick-Hall et al. This knowledge is crucial for effectively solving problems, such as those faced by train traffic controllers when confronted with multiple disruptions to the train schedule. Visual attention and evaluation of relevant information in complex problem situations improve when mental models are well developed. The degree of development of mental models differs between novices and experts.

Novices are therefore oriented toward surface characteristics in problem solving Bogard et al. Furthermore, experts have developed a condition—action ability through practice.

Experts have conditioned knowledge: The recognition of specific patterns triggers an appropriate response that is useful for problem solving Bransford et al. Different levels of expertise may influence the performance of train traffic controllers and therefore resilience at an individual level Lengick-Hall et al. The simulator was strongly focused on the logistical aspects of train traffic control and much less on technical safety-related aspects.

The infrastructure that was simulated was the train traffic area around Utrecht Central Station. This area is operated by two train traffic workstations. The role allocation was reversed in the second round. Two scenarios were designed for the participants: Scenario 1 consisted of a light disruption in the train traffic flow caused by minor train delays, whereas Scenario 2 represented a moderately to severely disrupted flow.

In the first round, train traffic controllers participated in Scenario 1. In the second round, 10 participants participated in Scenario 1 and 12 participated in Scenario 2. Both scenarios were designed in collaboration with two senior train traffic controllers.

Train traffic controllers were asked to perform their job as they typically would at their actual workstation. No interaction between the train traffic controllers was needed to conduct their tasks. All 22 train traffic controllers 18 males, four females worked at Utrecht Central Station. Work experience and job role were assessed using questionnaires.

Participants were assigned to a high- or a low-experience group based on their work experience as train traffic controllers.

The cutoff point was set at 10 years, as a new traffic management system had been implemented 10 years earlier Bary, A list of performance indicators for train traffic controllers was created prior to this session by six senior train traffic controllers. Speed of acquaintance was included to find out how fast the participants were able to get accustomed to the new infrastructure. This item was measured on a 5-point Likert scale, ranging from fully disagree to fully agree.

Participants could also opt for I do not know as an answer. Performance was measured using five performance indicators, namely, arrival punctuality, departure punctuality, amount of arrival delay, amount of departure delay, and platform consistency. Arrival and departure punctuality was operationalized through trains that arrive at or depart from Utrecht Central Station on time or with less than a 3-min delay. With regard to platform consistency, all trains that did not arrive at the planned track were counted and summed up, and the same was done for all trains that did not arrive at the planned platform.

Second, the total number of trains that did not arrive at the planned platform and at the planned track were summed up and divided by the total number of arrived trains for each train traffic controller. Simulator logs were used to analyze the completion strategies when different ways of dealing with the train delays i. Given the length of Scenario 1, three conflict points for completion strategies for the through workstation and one conflict point for the turn workstation were identified; for Scenario 2, one and two completion strategies were identified for the through workstation and the turn workstation, respectively.

Different completion strategies were subsequently assessed by analyzing whether the completion strategies were followed according to the preferred completion strategy as was scheduled and the different strategies applied, to assess the variability per operator and per conflict point.

Analyses were done based on participants who enacted Scenario 1 in both rounds and those who enacted Scenario 1 and subsequently Scenario 2, in order to obtain four conflict points per individual.

Simulator validity was measured through three components—structural validity the degree of similarity in structure between the simulated and the reference system , processes validity the degree of similarity in processes between the simulated and the reference system , and psychological reality the degree to which the participants perceived the simulated system as realistic —in line with Raser , using a questionnaire designed by Lo, Sehic, and Meijer The participants completed a questionnaire before the start of the session.

They then enacted the two min scenarios. At the end of each round, they completed another questionnaire. During the second round, knowledge probes were administered for the purpose of another study. Video recordings were made throughout both sessions. Six of the 22 participants were excluded from the simulator data analysis because they had known about the train delays. Another two participants were excluded as they enacted Scenario 2 twice. As there were a few problems with the simulator, not all train traffic controllers received the same number of trains.

This issue was controlled for by using an average score of the objective performance measures and reviewing the severity of issues through video recordings for events that hindered participants in their options or decision making. The findings show that the participants tended to be slightly positive about the validity of the simulator considering the task they were given see Table 1.

The participants also indicated that they had quickly gotten used to the simulator. Regarding learning effects between scenarios, the participants indicated that they had gotten used to both workstations relatively quickly. Figure 1 shows a relative moderate goal consistency among the train traffic controllers.

Three controllers added two more performance indicators, but they were not included in the analysis. As such, these results show a very fragmented preference with regard to primary key performance indicators.

Applied completion strategies per participant for operators who enacted Scenarios 1 and 2. A white band indicates a preferred completion strategy being followed, and a gray band indicates alternative completion strategies. Even numbers represent participants from the through workstation; odd numbers, those from the turn workstation. An analysis of the level of variation in completion strategies for each conflict point revealed diversity based on between one and three different completion strategies for four conflicting points in Scenario 1 and on five different variations of completion strategies for three conflicting points in Scenario 2 see Figure 3.

A qualitative assessment would show that there is a level of variation in the completion strategies with regard to different conflict points and that this variation differs between scenarios: Operators dealt with these conflict points with more diverse completion strategies in the moderately disrupted scenario than in the lightly disrupted scenario.

Further, it is notable that preferred completion strategies were implemented more frequently in Scenario 1. A white color indicates a preferred completion strategy being followed, whereas different shades of gray indicate different completion strategies. Numbers 1, 2, 5, and 6 represent conflict points from the through workstation, and numbers 3, 4, and 7 represent conflict points for the turn workstation. Spearman correlation tests were performed to test whether there is a congruence between the self-reported relative importance of performance indicators and objective performance see Table 2.

Although Scenario 1 does not reveal any significant correlations, Scenario 2 does, namely, a strong positive correlation between self-reported departure punctuality and objective arrival delay. Also a strong negative correlation was found between self-reported arrival punctuality and objective departure delay.

A trend for a negative correlation between self-reported platform consistency and objective arrival delay was also found. Although unexpected, these results provide interesting insights into goal competition, as they suggest that arrival punctuality and departure delay, departure punctuality and arrival delay, and platform consistency and arrival delay are competing goals.

A Spearman correlation test was also performed between the applied preferred and alternative completion strategies and performance. No significant relations were found. It was expected that the more experienced controllers would outperform the less experienced controllers due to their better organized mental models.

The analyses showed a significant tendency in Scenario 1 for controllers with less experience in their current function to have a higher arrival punctuality score than the more experienced controllers see Table 3.


Individual Markers of Resilience in Train Traffic Control



ProRail BV



Railway stations in the Netherlands




Related Articles