R2.119 Establishing Baseline Rail Level Crossing Incident Occurrence and Behaviours Using Video Data

A key to improving long-term safety at level crossings is to understand the factors that cause both accidents and near-misses.
Data on near-misses has previously not been consistently and centrally located. It has therefore not been possible to develop a holistic cause and effect model to reduce near-miss occurrences.
The purpose of this project is to address these issues and provide a framework for collecting and analysing data for the future. The project aims to:

• Create a system that collects relevant data on near-misses at crossings that will form the basis for the causation for near misses. Data sources include but are not limited to; cab video, cab GPS, ATP logs, signalling logs, and geo referencing.
• Use data to develop and refine a near-miss causation model. The final model will provide a clear and objective model for evaluating near-misses and improving safety at level crossings.
• Reduce the subjectivity of near-miss incident reporting through the use of a sophisticated model and provide consistent and accurate data for evaluation.

Project Status

The near-miss causation model is currently in development and the first version will be available late 2012. The input for the model has been provided through workshops with a number of industry experts. Design of the data capture and analysis system has commenced and is expected to be completed on schedule.


Implementation of a predictive model will improve capacity to prevent near-misses and accidents at level crossings as well. This will improve the safety of the networks and reduce the cost associated with accidents at level crossings.