Human perceptions of vehicle turning intention: overall performance and contributory factors

Abstract

All pedestrians, drivers and cyclists regularly make predictions on where they think an oncoming vehicle is intending to travel, so that they can successfully and safely navigate road systems. Despite the importance of these predictions, the effectiveness of this process is currently poorly understood with all existing research being focused on predictions from in-vehicle technologies. This paper therefore investigates how well observers are able to predict a vehicle’s turning intention as it approaches an intersection and explores the explanatory variables involved in the success of this process through a logistic regression analysis. An interactive touch screen experiment was developed so that people’s predictions about turning intention could be investigated. The data set has been created with over 100 participants attempting to predict a number of vehicles’ turning intention.
The key findings of this study are that people are very good overall at predicting turning intention with approximately 90% median success rate when vehicles are between 0 and 20 meters (0-21.9 yards) away from the intersection, but with a substantial fall to approximately 70% median success rate when the vehicle is between 30 and 50 meters (32.8-54.7 yards) away. Other key explanatory variables include both vehicle specific factors (e.g. use of indicator lights) and crucially the intersection layout, providing valuable information on the relationship between intersection design and road safety

More information

Identifiers

ORCID for Ben Waterson: orcid.org/0000-0001-9817-7119

Catalogue record

Export record

Altmetrics