A WHOLE OF LIFE CYCLE HUMAN FACTORS APPROACH
An important feature of the research programme was the use of human factors analysis and design approaches throughout the rail level crossing design life cycle. The majority of the research focussed on the early aspects of the life cycle; that is, methods such as Cognitive Work Analysis (CWA) and Hierarchical Task Analysis (HTA) were used to analyse existing rail level crossing environments (Chapter 5), to inform the design of new rail level crossing environments (Chapter 6) and to assess and refine the resulting rail level crossing design concepts (Chapters 7 and 8). Importantly, within these activities, issues across the system life cycle (i.e. implementation, maintenance, upgrades, decommissioning) were considered. As one example, engagement with stakeholders during design refinement activities enabled the information to be gathered regarding maintenance costs and concerns in relation to new rail level crossing technologies and infrastructure.
This whole of life cycle human factors approach is something that is often urged in system design, but is not often achieved (Stanton et al. 2013b). As many issues emerge from unexpected interactions at the boundary between people and systems, the need to engage in an iterative design-test-redesign process is paramount. This research programme has demonstrated how this can be achieved with approaches such as CWA.
As discussed earlier, an increasing number of researchers are arguing for a systems approach to be taken when attempting to improve transportation safety (Cornelissen et al. 2015, Larsson et al. 2010, McClure et al. 2015, Salmon et al. 2012b, Salmon and Lenne 2015). As research in the transportation domain has predominantly adopted an engineering-based approach, it is important to clarify the contribution of human factors and systems thinking approaches over and above the engineering approach to facilitate further systems thinking applications.
In applying the whole of life cycle approach, the current research programme demonstrated the benefits of applying systems thinking approaches to the analysis and design of level crossing systems. By collecting comprehensive data regarding the behaviour of different users and integrating these data within appropriate systems analysis frameworks, a rich and detailed description of rail level crossing system behaviour was produced. This considered multiple forms of users, along with various systemic factors that influence behaviour. Notably, these factors included those related to multiple stakeholders: from road users, rail level crossing designers, road and rail operators, to government and the community at large. A major strength of the approach is that it promotes consideration of not only the physical environment of the crossing but also the wider environment within which rail level crossings are designed, operated, maintained and upgraded. As such, recommendations arising from the research addressed changes at higher levels of the system (e.g. risk assessment processes, incident reporting systems, design standards and guidelines) in addition to changes to the design of rail level crossing environments (e.g. warnings, signage). The breadth of the analyses conducted throughout the research programme is such that they could also inform other reforms designed to increase the reliability, efficiency and usability of rail level crossings.
Another notable strength of the approach adopted is the range of human factors concepts considered. Through applying in-depth analysis methods such as CWA, HTA and Systematic Human Error Reduction and Prediction Approach, various aspects of behaviour were considered, including decision-making, situation awareness, errors and constraints. This aspect of systems analysis methods offers a significant advantage over methodologies that focus explicitly on individual concepts (e.g. human error) in isolation.
Finally, the formative component of the analysis and design approach is worth mentioning. This enabled analysts to explore how activity within the system could be undertaken given design modifications, which in turn supported the identification of important design insights. Using standard normative or descriptive analysis approaches would not have facilitated this.
Some pertinent weaknesses of the approach should also be noted. The analysis and design process adopted incurred a significant level of resource usage, with the overall programme of research taking 5 years to complete. This could be improved through the introduction of dedicated software support for some of the analysis methods (e.g. the strategies analysis diagram) and the use of automated data collection and analysis techniques (e.g. auto-transcription). Overall, however, the utility of the outputs produced in this case justify the high level of resources invested.
