A validation study comparing performance in a low-fidelity train-driving simulator with actual train driving performance

Olsson, Niklas · 2023 · Crossref

DOI: 10.1016/j.trf.2023.07.007

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses a gap in railway research by validating the performance of a low-fidelity train-driving simulator against actual real-world driving performance. While simulator validation is common in the automotive sector, no prior studies had established the validity of train-driving simulators. The research aimed to determine if a physically low-fidelity but functionally high-fidelity simulator could predict real-world competence, thereby offering a cost-effective and safe method for assessing train drivers and evaluating training interventions. The study involved 34 train-driver students in Sweden, divided into two classes. Class A (2019–2020) underwent standard training, while Class B (2020–2021) experienced COVID-19 restrictions, including reduced internship time and distance learning. Participants completed a 45-minute simulator test comprising freight and passenger train scenarios with normal driving and complex special cases. Performance was measured by counting driving errors, categorized as rule-based, skill-based, or ineffective handling. These simulator results were compared with internship grades provided by experienced supervisors over an 11-week period, using a standardized checklist. Statistical analyses included Pearson correlations and ANOVA to assess validity and the impact of external factors. The results demonstrated a significant negative correlation between the number of simulator driving errors and internship grades ($r = -0.45, p < .05$), indicating that fewer simulator errors predicted higher real-world performance. This confirms the simulator’s relative validity. The study also found that COVID-19 restrictions negatively impacted Class B, who committed significantly more driving errors and received lower internship grades than Class A. Specifically, Class B made significantly more skill-based errors. The presence of a peer observer during the simulator test had no significant effect on performance. The findings conclude that low-fidelity simulators with high functional fidelity are valid tools for measuring real train-driving performance. This validation supports the use of such simulators in both research and industry. Researchers can utilize them to study training methods and driver behavior, while train operating companies can employ them to assess driver skills and identify those requiring additional training before accidents occur. The study highlights that functional fidelity is critical for transferability to real-world tasks, even when physical realism is limited.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-09; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• simulator training transfer
• simulator validity fidelity
• rail grade crossings

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Applied Guidance: countermeasure evaluation
• Methodological Resource: validation psychometrics, tool software