Using naturalistic data to measure the contributors to serious bicycle crashes

In view of the urgent need of a pragmatic system to reduce cycling crashes, this project aims to derive new road safety insights and actions from deep analysis of a naturalistic-cycling dataset from an Australia-first trial (Light Insights Trial – LiT). This dataset was developed in a world-largest trial of a smart bicycle light in Victoria with more than 800 cyclists over 12 months. This project will undertake in-depth analysis of this unique naturalistic dataset in an augmented Statistical and Machine Learning modelling framework, as well as gather and analyse additional field data from high-risk cycling locations. This innovative approach does not replicate any existing cycling safety initiatives in Australia or internationally.