Marking Rail Safety Week 2026, this online session aims to share information on railway crossing safety projects. The content is relevant for rail and road infrastructure owners and managers, rail operators, governments, researchers and members of the community.
The session includes presentations including Q&A from each of the following speakers:
When: Thursday 13 August 2026 12.30-2pm (AEST).
Microsoft Teams meeting
Join: https://teams.microsoft.com/meet/44765052005260?p=AFpvVLcIQp1pD4TH7S
Meeting ID: 447 650 520 052 60 | Passcode: 7hr34c36
Presentation 1: Australian Transport Safety Bureau – Investigation into a passenger train and truck collision at a Runcorn, QLD.
Presenter: Deborah Loats, Investigator, Australian Transport Safety Bureau.
Despite continued efforts to prevent collisions at railway crossings, the ATSB continues to investigate similar occurrences. ATSB Senior Transport Safety Investigator Deborah Loats will talk about an investigation she led into a railway crossing collision in Queensland between a passenger train and a B-double truck in 2025.
Deborah will also highlight the importance of treating any track obstruction as an emergency, and the learning opportunities available to all road users and the rail industry.
Presentation 2: Improving escape gate usability at railway crossings – A Human Factors approach.
Presenter: Heidi Mundhenke, Human Factors & Systems Integration Lead, Safety & People, Metro Trains Melbourne.
An issue was identified around 2018 in relation to members of the public struggling to understand how to open electromagnetic emergency escape gates at pedestrian railway crossings.
The Rail Crossing Human Factors Working Group initiated an investigation to characterise the issue and evaluate options for improvement. Following further consultation with Metro stakeholders, a new set of signs has been proposed following VR trials and is progressing towards on-site trials on the Metro network.
Presentation 3: Enhancing Road Safety Planning at Level Crossings using Predictive Analytics.
Presenter: Simon Mac, (Act) Senior Manager PNAV Data & Analytics.
A Transport for NSW (TfNSW) initiative delivered, to address level crossing (RLX) safety. RLX crashes are infrequent but severe, and risk comes from complex interactions between road and rail factors. ALCAM, the current approach, relies on site visits. Resource limitations mean that individual RLX are infrequently assessed and rural and remote crossings are particularly challenging.
This initiative aimed to improve risk identification for 1,300 RLX across NSW by integrating available statewide datasets and new behavioural insights from Connected Vehicle Data (CVD) into a predictive machine learning model. The model combined ALCAM, AusRAP, ONRSR, RAMS, weather and geospatial data, with CVD from Compass IoT, to reveal the factors most strongly associated with safety risk, even in remote areas.
The project confirmed that machine learning approaches can be applied to predicting risk, and that behavioural insights add significant predictive power. We then carried out the first desktop systemic risk assessment of all 1,300 level crossings across NSW. The initiative demonstrates road safety impact by deepening the evidence available to predict level crossing crash risk, enabling more targeted investment and planning, focusing resources on locations with the highest risk.