Train drivers play a crucial role in the railway system, with their performance significantly impacting accident rates and punctuality. This is particularly true during abnormal situations, such as signal failures or the presence of people on or near the tracks. In these circumstances, good driver performance enhances overall railway performance, leading to better punctuality, fewer accidents, and a reduction in collisions between trains and other vehicles, animals, or people.
One of the most critical skills for a train driver to minimize errors is situational awareness, a person’s ability to comprehend what is happening in a given situation, which is essential for responding effectively (Endsley, 1995). This skill becomes especially important in abnormal situations. A situationally unaware driver results from either a lack of expertise or failure of expertise. Lack of expertise means that the driver does not possess the necessary knowledge to build an accurate situational model, increasing the risk of misinterpreting a situation. Naturally, novice drivers, lacking experience in many situations, are more prone to making mistakes due to this knowledge gap.
Basic train driver education in Sweden includes around 20 weeks of on-the-job training. However, it is debatable whether this provides enough exposure to various situations for drivers to gain adequate experience. In response to this criticism, a low-fidelity train driving simulator has been introduced in the last decade to complement traditional training. This thesis explores how the simulator can enhance the internship experience and help produce more well-prepared train drivers.
Still, even experienced drivers can misinterpret situations (failures of expertise), often due to distractions from multiple information sources, which may direct their attention away from critical elements. A train driver must process information from outside the train (e.g., signals, marker boards, and other objects) and from within the cab to build an accurate situational model. Modern train cabins feature at least three screens, providing vital information about the train’s condition, energy-efficient driving, and, most importantly, automatic train protection (ATP) systems, which display details like, speed limits and signal status. This complexity increases when drivers simultaneously gather and process information from outside and inside the cab. Consequently, this thesis also examines how in-cab information systems influence driver attention and the possibility of developing accurate situational awareness.
The thesis is composed of four papers. Paper 1 used web-based questionnaires to assess how frequently 43 specific situations arise and are practiced during internships. Paper 2 involved a simulator experiment, investigating how simulator training impacts driver performance in a simulated test. Paper 3 examined the relationship between actual driver performance, based on supervisor evaluations over 11 weeks of internship, and performance during a 45-minute simulator test, measured by counting driving errors. Finally, Paper 4 explored how in-cab information systems affect driver attention, utilizing eye-tracking data from a simulator experiment.
The findings reveal that, apart from various shunting scenarios, trainee drivers must be exposed to more practice in abnormal situations. Furthermore, the study shows significant variability in trainees’ experiences, posing a challenge for educators. Paper 2 demonstrates the effectiveness of simulator training, primarily due to its ability to offer repeated practice. The ecological validity of the simulator is supported by a medium-strong correlation between internship performance and simulator test performance, as discussed in Paper 3. Paper 4 highlights that the most critical in-cab information system, the ATP, can divert attention away from external cues, particularly during speed reductions. Additionally, the new European ATP system, known as the European traffic management system (ERTMS), demands more driver focus compared to Sweden’s automatic train control (ATC) system.
Based on these findings, simulator training should be extensively used to provide trainee drivers with practical experience in handling abnormal situations, which real-world conditions may not offer frequently enough. To optimize this training, a simulator test should first be administered to identify which scenarios an individual driver lacks experience in and needs to practice more. Furthermore, visual behaviour training should aim to enhance drivers’ external focus, particularly in areas where people are likely to be present, such as level crossings and stations. This would improve their chance to react in time to prevent potential collisions. In conclusion, to minimize the risk of errors stemming from a misinterpretation of situations (situational unawareness), simulator training should concentrate on providing sufficient experience in abnormal conditions alongside focused visual behaviour training.