Edinburgh Napier University

  The SiN (Safety-in-numbers) effect has become prevalent within cycling research in recent years and describes the phenomenon whereby cycling accident risk may fall when the volume of cyclists increases. Such an effect is clearly attractive to policy makers and campaigners and efforts to increase cycling volumes often now actively assume that a risk reduction and hence potential accident cost saving per cyclist will follow. However, according to UK and Scottish research, the SiN effect can coexist with a decline in cycle safety (Aldred et al., 2017) and evidence for beneficial SiN effect may be highly location dependent. Additionally, cycling risks are significantly higher, per kilometre travelled, than both motorised and pedestrian travel. Therefore, risk factors need to be properly understood in order to mitigate them and the presence or absence of SiN effect should be assessed at a relatively small area level. Specifically a national SiN factor should not be globally applied, and factors derived from other countries should be treated with caution.

To understand and measure SiN cycling flow data at both local and link level is required and this may not be readily available. Strategic cycling models such as Cynemon (TfL, 2017), which is similar to those that have long been used to analyse highways transport issues, are currently not feasible for most authorities due to cost. Cyclist flow is strongly correlated with the number of cyclists’ crashes such that more cyclists result in more cyclist crashes (although risk may reduce if SiN effect is present). Cycling flow data enables researchers to unpack risk causation and to unravel cyclist related incidents disaggregate between explanatory factors and exposure.

This paper presents a methodology to estimate cyclist flow patterns by utilising recently developed open source analysis tools (Lovelace et al, 2017) and cycling routing engine applications (www.Cyclestreet.net) which were developed specifically for cyclists. This is illustrated using a case study of Edinburgh City. A combination of traditional (Census and Automatic Traffic Counts) and novel (OpenStreetMap) data was used to produce flow estimates at both link and meso-spatial area levels. The range of novel cycling flow data sources available are discussed, while the case study to be presented utilises census data due to its reliability and lack of third party dependence. The variation of SiN effect at a meso-level within Edinburgh City will then be presented and comparisons made to the global Scotland SiN effect and to comparator regions from the EU.

The purpose of this research is to provide transport planners and policy makers with a quantitative and visual analysis of cycling flow to better understand road safety, risk and the SiN effect at a local area level. If increased risk is not mitigated the magnitude of injury and subsequent public health burden may continue to increase. Poor safety perceptions and outcomes will deter transfer to active modes of transport and future government investment may be threated if anticipated SiN effects fail to produce desired safety improvements.