Milestone’s video technology powers next-gen AI for smart cities

Open platform unlocks trusted video data and best-in-class video analytics for smarter, safer cities

Next generation AI for smart cities

The future of smart city technology isn’t being shaped in Silicon Valley–it’s unfolding in Dubuque, Iowa. This mid-sized city has become a live testbed for AI-driven traffic management through Project Hafnia, a pioneering public-private initiative led by Milestone Systems. The project showcases how responsible AI can transform mobility and safety–without costly infrastructure overhauls.

The challenge: high-potential AI, low-quality data

Many cities invest in surveillance infrastructure but fail to extract meaningful, real-time insights. AI models trained on synthetic or generic datasets often perform poorly in real-world environments, where high false-positive rates can undermine trust and usability in high-stakes scenarios like traffic management.

A collaborative solution

Led by Milestone Systems in partnership with the City of Dubuque and AI firm Vaidio, Project Hafnia developed a high-performance, real-world AI model. This 12-month initiative broke away from the traditional vendor-customer model. Instead, it was a true collaboration–each partner contributed expertise, infrastructure, and shared goals.

A key innovation was Milestone’s investment in a secure, legally compliant data library. By professionally annotating the city’s traffic footage, Milestone created high-fidelity training material that dramatically increased model accuracy–from 80 percent to over 95 percent.

Trusted AI at scale

Over 100 traffic cameras participated in the trial, testing AI performance across shifting light, weather, and traffic conditions. The result: a scalable, transferable solution that cities worldwide can adapt–not just for traffic monitoring, but for public safety, emergency response, and infrastructure planning.

“What makes this project unique is the approach to making the most of the data that cities already hold” said Ed Mauser, Director and Product Lead, Project Hafnia at Milestone Systems. “This public-private innovation is built upon thousands of hours annotating real-world footage so that computer vision models based on responsibly-sourced data in a compliant way.”

A blueprint for public-private innovation

Project Hafnia reframes how cities and tech companies work together. Dubuque, Vaidio, traffic experts, and Milestone, all had a hand in shaping the outcome and co-developing a flexible framework grounded in mutual value.

“Working closely with Milestone and Dubuque allowed us to fine-tune our models to real challenges,” said David Jenkins, VP of Software Architecture at Vaidio. “Training on real-world video data, not simulations, made all the difference.”

Infrastructure built for AI

Dubuque’s Traffic Engineering Manager, Dave Ness, has helped scale the city’s camera network from 20 to over 1,400. After early success with Hafnia, the city plans to extend the AI system across hundreds more intersections.

This reflects a shared strategy: build capabilities incrementally, not disruptively. Smart infrastructure doesn’t have to mean a complete overhaul–it can evolve through focused, collaborative innovation.

Responsible, compliant, future-proof

As AI regulation advances–like the EU AI Act–transparency and traceability are essential. Milestone created a fully documented data supply chain, with each frame of video traceable by source, history, and permissions. It complies with data protection standards like GDPR while establishing a foundation for ethical AI development.

Thomas Jensen, CEO of Milestone Systems, emphasised the broader vision: “AI innovation starts with trusted data and open collaboration. Project Hafnia proves even mid-sized cities can lead the way in responsible, data-driven transformation.”

A timely conversation among city leaders

Project Hafnia was recently presented by Ed Mauser and Dave Ness at the Cities Today Institute Leadership Forum in New Orleans, where city leaders gathered to explore the challenges and opportunities of municipal data licensing. The case sparked a lively, facilitated discussion on how cities can retain control over data while enabling innovation through responsible AI partnerships.

Looking ahead

Project Hafnia is more than a traffic management win. It’s a living prototype for ethical, scalable AI deployment in urban environments.

“We’re just scratching the surface,” added Ness. “This gives us insights and capabilities traditional systems couldn’t. It’s making Dubuque safer and smarter–for everyone.”

Want to unlock seamless, compliant, and high-quality video data for your AI models? Sign up as an early adopter and explore Project Hafnia before it becomes publicly available:

https://hafnia.milestonesys.com/joinwaitlist