TfNSW uses AI to predict accidents before they occur

Transport for NSW has identified previously unknown road black spots by using artificial intelligence and big data to analyse driver behaviour.

Juliana Bodzan

A pilot program carried out in Wollongong by the enterprise data and analytic services team was able to identify five dangerous intersections that weren’t previously flagged as black spots by analysing driver behaviour.

The 2019 trial, the results of which were announced last week, compared telematics data from the trial sites to data from intersections where accidents had previously occurred.

The telematics data identified behaviour like speed, harsh braking, sudden acceleration and lateral movement just before the intersection.

The fifty vehicles in sent 25 telematics records every second over 10 months, generating a massive amount of data.

Using a model developed from crash investigation data from other known black spots, the team was able to predict if an intersection was potentially hazardous.

“We had a circle of interest around the intersection,” data discovery leader at Transport for NSW Juliana Bodzan said.

“When is the vehicle actually approaching the intersection, how does it behave at 50 metres, how does it behave at 25 metres, how does it behave going through the intersection?”

Ms Bodzan says the department has been building its AI and machine learning capability since 2018 as part of its strategy to reduce road deaths and injury.

AI technologies provided by Microsoft including Stack, Databricks and Azure were used to ingest, curate and interpret the data.

Two of the five intersections identified during the proof of concept have now been targeted for safety upgrades in the next year.

The technology can also potentially be applied to railway lines and school zones, Ms Bodzan says.

“Transport is not just roads. We look at railway lines, we are thinking about pedestrians and school zones.

“And we are thinking about how we can use this type of data and enrich with other types of data to basically improve the safety of the network.”

Comment below to have your say on this story.

If you have a news story or tip-off, get in touch at editorial@governmentnews.com.au.  

Sign up to the Government News newsletter

Leave a comment:

Your email address will not be published. All fields are required