University of Melbourne and Cubic test AI camera to improve road safety

The camera can identify different road users as they pass into and through intersections.
Written by Aimee Chanthadavong, Contributor

The University of Melbourne and Cubic Transportation Systems have partnered to test how an artificial intelligence camera can be used to improve road user safety and traffic management.

As part of the project, seven cameras will be installed at various intersections along Melbourne's Rathdowne Street, which is part of the Australian Integrated Multimodal Ecosystem (AIMES).

AIMES, which is led by the University of Melbourne, was established in 2016 to live-test various transport technology on the streets of Melbourne in a bid to deliver safer and sustainable urban transport outcomes.

The Gridsmart cameras, develop by Cubic, uses real-time computer vision to track and identify between different road users -- including cars, motorbikes, cyclists, and pedestrians -- while they pass into and through intersections.

According to AIMES director at the University of Melbourne Majid Sarvi, being able to detect the different road users, especially "vulnerable" road users such as cyclists, motorcyclists, and pedestrians, will help determine how much green time individuals would need to cross an intersection safely.

"Our current transport system can only detect cars or trucks at one particular location, which is right behind the stop line. That was developed in 1979 or 1980 … and it has its limitation," he said.

"For example, you don't know the length of the queue behind the intersection, how many cars are waiting, how many bicycles are waiting, or how many pedestrians are waiting on the sidewalk to cross. If you don't know, you cannot manage it.

"With this camera technology, all of this is possible. The good thing is they can analyse the whole traffic using the processing embedded into them, and you don't even need to lift the image from the camera."

See also: 3 necessary components of successful smart cities (TechRepublic)

Image: Cubic Transportation Systems

Sarvi said the plan is to also use the camera, for instance, to be used for detecting near-miss incidents.

"If we can collect data about those incidents, then we can think about what we can do now, can we change the signal setting somehow to reduce the risk of crashes … that's a perfect way to look at what is happening," he said.

Set to go live in September, the cameras, according to Sarvi, will complement the existing 250-plus sensors that currently exist within the AIMES.

He said the cameras will be part of ongoing work by the university to "collect the right data, make sure it's correct, then look at how to use it, how to show the effectiveness of it", with hopes the data will be used to inform how the Victorian Department of Transport manages traffic.

Earlier this year, Cubic teamed up with Moovit as part of a definitive agreement to jointly develop a mobile solution for public transit agencies.

Under the partnership, Cubic will integrate Moovit's Mobility-as-a-Service (MaaS) solution with its own mobile Travelers app to deliver service alerts, information about nearby transport service lines, multimodal trip planning capabilities, and real-time arrival information. Ticketing for a journey and transit options that can be filtered based on cost or trip duration will also be integrated, the companies said. 

When completed, the platform will be made available to existing Cubic's customers, including public transit agencies in Brisbane, Boston, Chicago, Los Angeles, Miami, New York, San Francisco, and Washington, DC.

Cubic has been previously charged with developing Queensland's public transport ticketing system after the state government earmarked AU$371 million for the upgrade as part of its 2018-19 Budget.

Related Coverage

Editorial standards