ERTMS transitions: what impact do they have on train drivers?

During European Rail Traffic Management System (ERTMS) rollout in Great Britain, train drivers will be exposed to transitions between conventional signalling (with Class B systems) and ERTMS. On the approach to a transition... [ view full abstract ]

During European Rail Traffic Management System (ERTMS) rollout in Great Britain, train drivers will be exposed to transitions between conventional signalling (with Class B systems) and ERTMS. On the approach to a transition border, the driver will receive an announcement and is required to prepare for the new method of operation. The driver must acknowledge the transition at the transition border within five seconds of the acknowledge indication appearing on the driver machine interface (DMI). These tasks are in addition to the normal driving task. After the transition, there may be an adjustment period until the driver is secure in their new method of operation which may lead to degradation in the effectiveness of the driving task. The objective of this research was to identify safety and performance issues that could be attributed to transitions between ERTMS Level 2 and conventional signalling systems. There were five work packages completed: literature review, interviews with train drivers who currently transition between signalling systems, risk analysis workshops with drivers and operations standards managers from passenger and freight operators, a rail action reliability assessment (RARA) to estimate human error probabilities (HEPs) at key infrastructure features and analysis of data contained within the RSSB Safety Risk Model (SRM) (version 8.1) to estimate the impact of transitions on fatality and weighted injuries (FWI). There was evidence of distrust of new signalling systems when first introduced but also for prolonged periods (years) after implementation. This was particularly evidenced in experienced drivers. The distrust stemmed from the fact that the new signalling systems were perceived to be more supervisory and therefore de-skilling the role of a train driver and technical failures had affected trust in the system. During introduction of the new signalling system, drivers found that their workload increased and they were more likely to get distracted by in-cab indications. The transition activities became more automatic with time and attention could be switched between in-cab indications and the visual scene outside the cab. However, increasing the number of transitions, degraded operations or introducing transitions at high complex locations may introduce risks related to high workload and confusion. The HEP figures from the RARA assessment showed that performance tended to be at the more reliable end of the human performance spectrum. This was because the tasks effected by transitions were highly practiced and familiar but did identify errors with higher likelihoods. The analysis of SRMv8.1 data showed that increases in FWI due to the presence of transitions at key infrastructure features are very small compared with overall risk figures in the RSSB safety risk model. However, the analysis did show that locating transitions on plain line sections of track is preferable compared to junctions and speed restrictions. The limitations of this data should be considered when interpreting these results. The output of this project was a process for evaluating the design of transitions based on the Taking Safe Decisions risk management framework (RSSB, 2014). This process considers the human factors impacts of the transition on the driver.