Can AI solve Australia’s traffic gridlock?

Brisbane was ranked as the 12^th worst city in the world in terms of traffic congestion by the 2023 Global Traffic Scorecard. Over 1,200 road fatalities were reported in 2023 – the highest in more than a decade. We need a solution, and we need it now.

Australia’s booming population and the love affairs with cars are intensifying the traffic woes in major cities.

Australia embraces new technology for road safety

Outdated technology is holding us back. Most of our current systems still depend on human oversight and traditional methods which struggle to keep pace with the nation’s growing population and increasing traffic volumes.

Recognising this challenge, the Australian Government has committed to a “safe system” approach and the “Vision Zero” initiative, aiming to eliminate road fatalities and serious injuries by 2050.

Artificial intelligence (AI) can play a crucial role achieving this ambitious goal by acting as a powerful catalyst for a paradigm shift in traffic management.

Here’s how AI can enhance traffic management systems and potentially bring “Vision Zero” closer to reality:

Smart traffic counting: Precision in motion

The sheer volume of data generated by Australia’s growing fleet of vehicles is overwhelming existing systems. The influx of data leads to inefficiencies and missed opportunities to improve our traffic condition.

Vision AI can substantially reduce traffic accidents and congestion. This cutting-edge technology uses machine learning and neural networks to turn visual data into actionable insights.

It can accurately identify, count and classify vehicles, giving us a clear picture of traffic flow. This data empowers us to take proactive measures to prevent accidents and alleviate congestion.

A notable example comes from Montreal, which historically struggled with severe traffic congestion, particularly around its port. The city has deployed Fujitsu’s AI-enabled data analysis platform to analyse the traffic flow of around 2,500 traffic lights. This solution enables the city to make data-driven decisions to streamline traffic flows, lower fuel consumption and reduce air pollution.

Swift incident detection: Responding at lightning speed

In a traffic emergency, every second counts. Traditional sensors are often too slow to react, leaving us vulnerable to accidents and gridlock.

AI can enhance these systems with faster, more accurate algorithms that automatically detect incidents and notify road monitoring operators.

Mahesh Krishnan. Image: Fujitsu Australia

Japan’s traffic management systems is already reaping the rewards. Fujitsu’s AI-powered traffic management systems have revolutionised highway safety, especially during severe weather. By providing real-time traffic condition data, this technology has enabled traffic departments to pre-emptively prevent potential accidents, demonstrating the effectiveness of AI in real-time incident management.

Advanced traffic insights: See what you can’t see

Traditional traffic management systems struggle to analyse complex traffic conditions in real-time, particularly in areas with limited visibility or “blind spots.”

AI technology has the potential to address this issue through its advanced modelling capabilities, allowing traffic systems to “see” through obstructions.

Fujitsu and Carnegie Mellon University in Pennsylvania, USA, have developed AI-powered social digital twin technology with data from Pittsburgh. The newly deployed AI system was able to capture more detailed real-time traffic images and identify potential dangers in blind spots, significantly reducing the rate of traffic accidents.

Implementing AI requires a strategic approach

To successfully integrate AI, organisations need to establish robust data collection infrastructure, scalable cloud architectures and effective alerting mechanisms to support real-time decision-making.

Robust data collection: We need high-resolution cameras and sensors strategically placed across our road networks to collect comprehensive data.
Scalable cloud architecture: As the volume of camera data increases, we need a scalable cloud architecture to handle the processing power required.
Effective alerting mechanisms: AI insights must seamlessly integrate with existing traffic control systems such as traffic lights and digital signage for real-time action.

Integrating AI-powered computer vision into traffic management systems has the potential to revolutionise urban mobility in Australia. This technology will become even more crucial to ensure safe transportation, as we move into an era of self-driving vehicles.

By harnessing real-time data analysis, we can create a more efficient, responsive, and scalable solution that addresses the complex challenges of modern traffic management.