Cyclists' perceived safety on intersections and roundabouts - A qualitative bicycle simulator study

Introduction: Although cycling provides both individual and societal benefits, the mode share in Germany remains at a relatively low level. One reason described in literature is the lack of perceived safety due to the cycling infrastructure, especially at junctions. The study addresses the influence of junction design on cyclists' perceived safety.

Method: Three intersections (BS: Berlin Standard, PI: protected intersection, CbC: cycle lanes between car lanes) and one roundabout were modeled in a virtual environment. Using a bicycle simulator, n = 46 participants cycled through each junction design, followed by a qualitative interview. We conducted a structured content analysis on the interview transcripts.

Results: Regarding the quality of statements, PI provides the highest level of perceived safety whereas CbC provides the lowest level. Both roundabout and BS provide medium to low perceived safety. Specific design features, such as continuous cycling infrastructure, physical separation and elements enhancing cyclists' visibility improve participants' perceived safety. On the other hand, curbs, bends, and elements obstructing visibility decrease perceived safety. Our findings also point towards a difference between overextending and manageable interactions between cars and cyclists. While manageable interactions raise attention to an appropriate extent, overextending interactions diminish the quality of the cycling experience so that some cyclists rather violate rules instead of using the designated cycling infrastructure. Furthermore, three factors influence participants' perception of infrastructure design: comprehensibility, comfort, and perceived safety.

Conclusions: To provide a cycling friendly infrastructure, planners should consider cyclists' perceived safety as well as comfort and comprehensibility. Furthermore, in contrast to isolated segments, a continuous high-quality cycling infrastructure network should be implemented. Lastly, infrastructure might focus on manageable interactions rather than cause overextending interactions.

Practical application: The findings should be considered in future cycling infrastructure planning. Planners may test and modify temporary solutions to find appropriate designs for each junction.