What factors influence risk at rail level crossings? A systematic review and synthesis of findings using systems thinking

• Published in: Safety science 2021-06, Vol.138, p.105207, Article 105207
• Main Authors: Read, Gemma J.M., Cox, Jolene A., Hulme, Adam, Naweed, Anjum, Salmon, Paul M.
• Format: Article
• Language: English
• Subjects: AcciMap; Crashes; Decision making; Level crossings; Pedestrians; Rail level crossing; Railroad accidents & safety; Railroad crossings; Railway crossing; Risk perception; Road user behaviour; Roads; Safety; Systematic review; Systems analysis; Trains
• ISSN: 0925-7535; 1879-1042
• DOI: 10.1016/j.ssci.2021.105207

•88 studies relating to risk at rail level crossing were reviewed.•The influencing factors identified were mapped using a systems thinking framework.•The findings suggest little progress in implementing systems thinking approaches.•A ‘many models’ research agenda is proposed. Collisions between road users (drivers, cyclists, motorcyclists, pedestrians) and trains at rail level crossings (RLXs) remain an intractable issue. This systematic review aimed to determine what is known regarding the factors influencing risk at RLXs, with a focus on understanding the causal relationships across the entire sociotechnical system. A systematic search identified 88 published studies. The studies were categorised according to the type of outcome measured or analysed: (1) rates and severity of crashes; (2) unsafe and non-compliant road user behaviours; and (3) road user risk perception, attitudes and beliefs. Most studies focused on unsafe and/or non-compliant road user behaviour. The factors identified within the studies as influencing risk at RLXs were classified using the Accident Mapping (AcciMap) technique, a systems analysis framework. AcciMap maps the factors that influence behaviour across six system levels ranging from Government to the operating environment. Most of the factors identified within the studies related to physical attributes of the crossing itself, its operation, and the behaviours and attributes of road users. Comparatively fewer systemic factors were identified (i.e. policy and budgeting). Few relationships between factors were identified, indicating that little consideration has been given interactions between factors (i.e. how crossing design influences end-user decision making). A research agenda is proposed based on systems thinking and the use of a ‘many model’ approach to understand and address risk at RLXs.