The systems approach adopted ensured that various recommendations for improving safety at rail level crossings emerged. The recommendations are summarised using Rasmussen’s (1997) risk management framework in Figure 11.1.

The recommendations presented in Figure 11.1 can be broadly categorised as follows:

  • 1. Recommendations for the development of in-vehicle devices
  • 2. Recommendations for changes to infrastructure at urban rail level crossings
  • 3. Recommendations for changes to infrastructure at rural rail level crossings
  • 4. Recommendations for improving safety management around rail level crossings generally

Importantly, although a number of recommendations are focussed on changing the physical environment of crossings, the research programme produced recommendations that span the entire system.


Factors Influencing User Behaviour

In Phase 1 of the project, described in Chapter 4, we used human factors methods to collect a wide range of data about human behaviour at rail level crossings. This represented a significant research effort in its own right and highlighted some important findings. For example, we found differences in how novice versus experienced drivers interact with rail level crossings, with novices expecting active controls at all crossings (an expectancy violated at passive crossings) and failing to consider the possibility of queuing at busy urban crossings. Further, the findings highlighted differences in information use across different types of road user.

However, using these data to generate the systems descriptions of rail level crossing with CWA and HTA provided insights over and above those generated from the initial analyses. For example, the Work Domain Analysis (WDA) presented in Chapter 5 provided an actor- and event-independent description of rail level crossing environments, identifying the various physical and abstract constraints on their functioning. This provided crucial insights into the wider systemic issues that drive the configuration of rail level crossing environments, and thus influence road user behaviour. The systemic factors identified via the WDA included the underlying assumption of the system that road users must give way to trains, the competing

Recommendations for improving rail level crossing safety. GPS, global positioning system; ITS, intelligent transport system; RLX, rail level crossing

FIGURE 11.1 Recommendations for improving rail level crossing safety. GPS, global positioning system; ITS, intelligent transport system; RLX, rail level crossing.

pressures around safety and network efficiency and the focus on conformity with standards in system design and redesign.

It is worth noting that, for the research team, the WDA provided an in-depth overview of the functioning of rail level crossing systems and a way to identify potential conflicts and issues across the system. The WDA representation also resonated well with stakeholders as it provided a tangible way to see the how elements in the system were connected, in a way not previously available.

The relationship between the environment and the behaviour was then emphasised through application of the strategies analysis diagram, which explicitly links behaviours to the physical features identified in the WDA. This provided insights into how the underlying structure of the rail level crossing system ultimately influences user behaviour.

Related to this, another important outcome of the systems modelling was a detailed understanding of how different types of road users make decisions to stop or proceed (decision ladders) and the strategies they use to traverse crossings (strategies analysis diagram). The systems analyses enabled us to clearly identify where different road user groups adopt similar or consistent strategies, as well as where conflicts may arise. In short, we found considerable variation in how road users interact with rail level crossings and a range of ways in which adverse events can emerge. This raised implications for both design and redesign; for example, it was vital to be cognisant that design solutions aimed at car drivers would not have the same impact on heavy vehicle drivers or pedestrians. This was again emphasised in the differential ratings of novel rail level crossing designs provided by different road user types (Chapter 10). The importance of considering all road users in analysis and design is thus emphasised, as is the need to continually monitor and assess system behaviour and the unique challenges faced by different users.

Finally, the contextual activity template (CAT) and the Social Organisation and Cooperation Analysis (SOCA) prompted us to take very different perspectives on the system, compared with traditional human factors analyses that focus on tasks and activities. The CAT revealed the situational constraints on how the system can function and highlighted a range of possibilities for improving system design through modifications to constraints. For example, the CAT showed that information about risk is not currently provided to road users as they approach and traverse rail level crossings. This raised the idea of providing this information on approach to individual crossings, rather than holding it in documents and databases not accessible to users. The SOCA allowed us to examine the roles that human and non-human actors currently play in rail level crossing system operation, along with the roles that they could potentially play given design modifications. This enabled identification of insights that supported the generation of new design ideas. For example, when examining the function ‘Alert user to presence of rail level crossing’, the SOCA identified that actors outside of the road and rail infrastructure could potentially fulfil this function. This included the potential for using an in-vehicle display to warn of an upcoming rail level crossing. This potential was realised in both the Intelligent Level Crossing and GPS Average Speed design concepts.

Overall, the argument that systems analysis provides considerable insight beyond traditional human factors studies has implications for the design of future human factors research programmes. Given the acceptance within safety science that transportation systems are complex sociotechnical systems, this raises the question of how we can better support research that is integrated within a wider systems thinking framework. Acknowledging that many research endeavours do not have the time, resources or buy-in from funding agencies required to undertake in-depth systems analysis, there is a need to encourage future applications where systems models can be shared across research teams. This can avoid reinforcing system design recommendations that cater for some user types to the detriment of others, or that introduce the components into the system that have unintended and unforeseen negative consequences. For example, a comprehensive systems model of a road transport system could be shared by research groups and used to determine the design of smaller studies with the findings ‘rolled up’ to continually refine the larger model. Such efforts may be facilitated by the current trends towards open science and new technologies for collaboration.

< Prev   CONTENTS   Source   Next >