On-train-data-recorders OTDR are increasingly equipped on modern trains to capture information on train operations including speed, distance travelled, faults recorded, and GPS coordinates. Their primary purposes are accident investigation and train fault management, however, the OTDR also capture a range of parameters that may be directly and indirectly related to driver performance. In particular, the OTDR can capture all control actions taken by the driver, including power and brake applications, use of horn and headlights, door operations, response to warnings, and, in some cases, the currently applicable signal aspect. Such data provides new opportunities to explore train driver strategies and behaviours.
The purpose of this paper is to provide an overview of the possible human factors applications of the OTDR, and give examples from initial analyses of how the data may be applied. The contributions of other researchers (e.g. Rashidy et al., 2016; Walker & Strathie, 2015; Green et al., 2011) working in the area will be discussed, and three main applications proposed; first, modelling of driver journey profiles to understand differences within and between drivers in terms of their use of the train controls. Example results in this area show the different strategies employed by drivers and the possible future applications in linking driver profiles to performance will be discussed. Machine learning is a promising approach in this area, providing the ability to identify clusters in the large data sets which can be interpreted in the context of driver performance. Second, driver taskload modelling can be performed using the OTDR data to identify and analyse discreet control actions taken by the driver. Thus, the data can be used to examine differences in driver workload over different routes and potentially identify periods of underload. Examples of taskload calculations for different routes will be given. Finally, where aspect data is available the OTDR, as it is on the Irish Rail network, it may be used to investigate the approaches to red signals made, driver responses to those signals, and hence improve the understanding of SPADs. Example data will be shown describing key metrics that may be taken from this data. The paper will conclude with a discussion of the most promising avenues for further research.
Walker, G., & Strathie, A. (2015) Leading indicators of operational risk on the railway: A novel use for underutilised data recordings. Safety Science, 74, 93-101.
Green, S.R., Barkby, S., Puttock, A., & Craggs, R. (2011). Automatically assessing driver performance using black box OTDR data. In Proceedings of the 5th IET Conference on Railway Condition Monitoring and Non-Destructive Testing (RCM 2011) (pp. 1-5). New York: Institute of Electrical and Electronics Engineers (IEEE).
Rashidy, E.L., Ahmed, R., & van Gulijk, C. (2016). Driver competence performance indicators using OTMR. In Proceedings of CIT2016 Congreso de Ingeniería del Transporte (XII Congress of Transport Engineering) (pp. 354-361). Madrid: Foro de Ingeniería del Transporte.
Train driving models and performance , Signals and signage; SPADs , Human error and human reliability