Five large cities in Noord-Brabant and the provincial authorities investigated how to improve cycling safety through behavioural interventions. They placed signs and markings along cycle routes in 's-Hertogenbosch, Breda, Eindhoven, Helmond and Tilburg. Half of the interventions had an effect; the number of dangerous situations decreased.

Why this experiment was conducted: redesigning road layouts is not always possible
Six locations in towns and cities in Noord-Brabant were identified as unsafe for cyclists. A disproportionally high number of accidents and near accidents occur at these spots. As there were no feasible or practical options to significantly alter the road layouts there (e.g. by adding traffic lights or speed humps), the Province of Noord-Brabant decided to explore potential behavioural interventions.

Types of intervention: signs and markings
It was decided to execute the following behavioural interventions at the six locations:
• a sign was put up at a set of traffic lights to prevent schoolchildren from cycling through red lights (Breda)
• signs were placed at the spot where a cycle through-route crosses a 50 km/h road, to get motorists to appreciate how cyclists perceive the road and traffic (Eindhoven)
• a silhouette of a cyclist was put up at the spot where a cycle through-route crosses a 50 km/h road, to make motorists aware of cyclists (Eindhoven)
• signs were erected along a cycle through-route, to encourage cyclists to make contact with motorists and reduce speed (Helmond)
• signs and markings were placed on the road to encourage cyclists to make hand signals at a roundabout ('s-Hertogenbosch)
• signs and markings were placed on the road to encourage cyclists to cross a 50 km/h road safely, by first stopping ('s-Hertogenbosch)

Method used: baseline and impact measurements and follow-up
A baseline measurement was performed at all the locations, followed by an impact measurement after 1-2 weeks, and a follow-up measurement after 2-3 months. At these locations, cyclists' and motorists' behaviour was regularly observed: did cyclists stop at red lights or before crossing, did they make hand signals at the roundabout, and/or did conflict situations occur between cyclists and motorists? A conflict refers to a situation where a motorist does not immediately give right of way to a cyclist, for example, if a cyclist brakes hard or too late, if a cyclist passes just in front of a car or vice versa, or if there is an accident, such as a collision or a fall.

Result obtained: half of interventions were effective
In the case of four of the six interventions, the incidence of observed unsafe conditions decreased significantly in the impact measurement. In the case of three of these interventions, that effect persisted longer, i.e. even during the follow-up two months later. The effect is especially large in the case of the signs that get motorists to appreciate how cyclists perceive the road and traffic. In Breda notably, the effect was eliminated during the follow-up measurement, even though the sign was still in place. It is possible that the effect has waned because pupils pass it daily on their way to and from school. In Tilburg, the study specifically examined situations involving an approaching car. The findings were: in the baseline measurement, 55% of cyclists stopped; in the impact measurement, 60% of cyclists stopped; and in the follow-up measurement, 76% of cyclists stopped. In addition, fewer near-accidents were observed: 10, 7 and 1 times, respectively.

Impact: promising interventions. including for other locations
The test provides insight into ways in which the Province can improve cycling safety. Also, more insight was gained into the long-term impacts and the duration of the effect of behavioural interventions. The signs and markings remained in place after the experiment ended. Three interventions displayed potential for application elsewhere, in particular the signs that get motorists to appreciate how cyclists perceive the road and traffic.

Source: https://www.binnl.nl/home+-+en/knowledge/publications/bin+nl+publications/HandlerDownloadFiles.ashx?idnv=2719979