AI for smart cities is no longer a futuristic concept. With the launch of Project Hafnia, Milestone Systems is offering a real-world example of how responsibly built computer vision systems can support more liveable, efficient cities — and it’s a framework that could readily benefit councils across Australia and New Zealand.
Project Hafnia, built in collaboration with NVIDIA, is already helping transform urban environments by providing regulation-ready datasets to train artificial intelligence systems. While the initial rollout is focused on European cities, the technology is highly adaptable — and ANZ city planners, transport authorities and infrastructure leaders should be taking note.
Using NVIDIA’s AI Blueprint and Omniverse platform, Hafnia allows cities to create digital twins — virtual replicas of real infrastructure — to test and improve traffic flow, public safety, and emergency response protocols. It’s a compelling solution for cities like Sydney, Melbourne, or Auckland, where urban congestion, infrastructure strain and sustainability goals are top of mind.
The AI models being developed are powered by vision language models (VLMs), capable of understanding video and image data in context — summarising scenes, detecting anomalies, and helping operators make faster, data-driven decisions.
What makes Project Hafnia stand out is its ethical backbone: all data is curated under strict privacy and compliance standards, creating a framework that mirrors Australia’s own growing push for responsible AI governance.
“Any city with a forward-looking digital strategy can leverage this platform,” said Jensen. “From real-time traffic analytics to predictive planning, Project Hafnia can help turn big video data into real, actionable outcomes.”
