An investigation into the relationship between turning signal and crash risk on urban roads

The relationship between the relative risk of a collision involving a turn maneuver and the configuration of both the rear and front turn signal (ON/OFF) was examined using data pooled from the answers of a survey proposed to 67 people. To perform the analyses, many videos have been collected during the tests campaign, both in urban and suburban areas, using a 360-camera attached to the motorcyclist’s helmet. In the videos the biker was riding along the street and was overtaking/passing a car stationary at an intersection which had the turn signal turned on. Different variations of this basic situation have been performed such as, adding a second car behind the first one, turning off the car’s blinker, changing the location of the tests (from urban to suburban area), changing the direction of the motorcycle (from having the same direction of the tester’s car to opposite direction). During the survey phase, people needed to answer a simple questionnaire- after having watched two videos (chosen among the twelve alternatives) using a virtual reality visor- containing, among the other questions, the one asking if they saw the car’s turn indicator turned on or not. The detection of the blinker was combined with other factors (e.g. age, gender, location, presence of the car behind the tester vehicle etc.) in a stepwise logistic regression that modelled the odds of detecting the turn signal turned on as a function of all of these factors.The analyses suggest that there is an association between detecting the turn signal switched on and, for example, being female, or not being a habitual biker, or being in suburban area. The differences in the analyses are discussed in terms of the attributes (age, gender, etc.) chosen for the different tests configurations. Further investigations of factors and data seem warranted before drawing more accurate conclusions regarding the turn signal and the association with the attributes of the model.