How integrated video analytics is helping cities tackle urban mobility

As Australia's major cities grapple with the dual challenges of population growth and ageing infrastructure, the promise of smarter, data-driven traffic management is no longer a futuristic ideal - it's a necessity. With congestion projected to cost the Australian economy over $39 billion a year by 2031and urban mobility becoming a flashpoint for local councils and state governments, the integration of video analytics into city management systems offers a compelling, cost-effective solution.

At the heart of this transformation is the convergence of artificial intelligence (AI) with video surveillance infrastructure, turning passive cameras into active sensors that can deliver real-time traffic intelligence. This integrated approach is being trialled and adopted around the world, including in mid-sized cities like Dubuque, Iowa, providing a blueprint for Australian councils seeking scalable, affordable solutions to transport-related headaches. 

When training an AI model to deliver real-time intelligence like this, high-quality data is essential. AI models require vast amounts of structured data to develop reliable insights. Milestone's Project Hafnia for example collects video data from various sources and curates it using NVIDIA NeMo Curator, ensuring high-quality and regulatory-compliant data for projects such as the city of Dubuque and others.

From surveillance to situational awareness

Traditionally, video surveillance in Australian cities has been used for security and enforcement - monitoring public spaces, deterring anti-social behaviour, or issuing fines for traffic violations. However, the emergence of AI-powered video analytics is radically expanding these use cases. Cameras are now capable of identifying stalled vehicles, detecting drivers going the wrong way, monitoring pedestrian flows, and even estimating CO₂ emissions based on traffic patterns.

When combined with modern video management systems (VMS), these analytics can feed into centralised dashboards, giving urban planners and transport authorities a real-time pulse of the city. The result is faster incident response, improved road safety, and more responsive traffic signal optimisation - without needing to dig up roads or deploy costly new sensors. The infrastructure is already there.

In Dubuque, for example, city officials integrated intelligent video analytics with their existing traffic camera network to detect incidents in real time and analyse patterns in road usage. This enabled data-driven decisions around peak congestion times, pedestrian safety zones, and vehicle prioritisation, outcomes that could translate seamlessly into the Australian context, particularly in regional centres and high-growth suburbs. 

For the big cities like Brisbane, Sydney, and Melbourne - where population growth continues to outpace infrastructure investment - solutions that leverage existing assets are critical. Integrated video analytics allow councils to repurpose existing CCTV infrastructure for mobility, but also to improve climate and safety outcomes.

In regional and suburban contexts, where budgets are tighter and transport planning cycles longer, the appeal is even clearer. Rather than installing a patchwork of Internet-of-Things (IoT) sensors, environmental monitors, and proprietary traffic systems, video-based analytics provide a unifying layer that is both hardware-agnostic and scalable.

This aligns well with the trend towards open platforms - systems that can integrate with a variety of third-party analytics tools and datasets, rather than locking cities into closed, siloed ecosystems. Open architecture is especially important in Australia, where federated systems of government often result in fragmented data ownership and technology procurement.

Privacy, policy and public trust

Of course, deploying smart cameras and AI analytics in public spaces raises important questions around privacy and governance. In Australia, local governments must navigate not only technical challenges, but also the regulatory and ethical frameworks that govern public surveillance and data use.

Success stories overseas suggest that transparency and community engagement are key. In Dubuque, public trust was built through open communication about how video data was being used, what was being recorded, and how long it would be retained. Applying similar principles locally will be essential to gaining community support for video-based mobility initiatives.

Crucially, the technology is advancing in ways that respect privacy by design. Many systems now allow for object-based detection (vehicles, pedestrians, bicycles) without facial recognition or identity tracking, making it possible to analyse urban movement patterns without compromising individual rights. This allows a council to stay well away from any risk associated with Australia's Privacy Act, and maintain compliance with regulations.

As Australia moves towards a future of connected, autonomous and electrified transport, the foundations for smarter mobility must be laid today. Integrated video analytics offer a practical, affordable and scalable way to start building that foundation, whether it's in dense city centres, regional hubs or rapidly growing outer suburbs.

While major infrastructure projects such as new rail lines, motorways and tunnels can take years to complete, upgrading a city's situational awareness through video intelligence can be done incrementally and with immediate impact. This makes it an ideal solution for councils looking to show quick wins while building towards longer-term transformation.

As seen in international examples like Dubuque, the results can be revolutionary. The challenge now is for Australian cities to take that first step and to turn the lens of surveillance into a tool for smarter, safer, and more efficient mobility.