Milestone unveils AI plug-in to cut false video alarms by 30%
Milestone Systems has announced a new generative AI-powered video analytics plug-in for its XProtect video management software, developed in collaboration with NVIDIA. The company says the tool is designed to help operators filter out false alarms, contextualise incidents more effectively, and reduce the burden of video review in high-volume environments such as urban centres and traffic management operations.
Automated analysis
The new XProtect plug-in leverages generative AI to automatically summarise, contextualise, and validate video content in real time. By doing so, it aims to help security teams and city operators handle incident response more rapidly and reduce the growing problem of operator alarm fatigue. According to preliminary findings from Milestone, this technology has the potential to decrease operator fatigue from alarms by up to 30 percent.
The plug-in includes automated incident report generation, which instantly converts selected video clips into incident summaries and structured reports. This approach is expected to reduce the time spent on manual documentation. Additionally, the software can be configured to analyse motion events and validate alarms through direct integration with the XProtect rule engine. This, according to the company, should help cut false positives and streamline alert management.
Another feature allows for automatic summarisation of bookmarked footage with natural language output, speeding up the triage process and making it unnecessary for operators to manually review each individual clip.
Deployment options
The plug-in is compatible for deployment both on-premises and in the cloud. This flexibility is suited for customers with strict compliance requirements or unique infrastructure needs. Direct integration with the XProtect rule engine is intended to assist organisations in refining workflow and policy enforcement without additional complexity.
Training and compliance
The generative AI model, known as Hafnia Vision Language Model, has been trained on 75,000 hours of ethically-sourced, real-world video data. All data originates from either Europe or the US and is prepared using the NVIDIA Cosmos Curator platform. The analytics engine uses NVIDIA infrastructure either on cloud or regional data centres. The joint use of NVIDIA Cosmos Reason VLM positions the solution as one of the more advanced and compliant offerings for video management analysis in the sector.
Industry application
Organisations including the cities of Genoa in Italy and Dubuque in Iowa have indicated an interest in deploying the new plug-in, highlighting growing municipal demand for AI-enabled video analytics to improve traffic system management and public safety operations.
Platform expansion
Apart from the plug-in for XProtect, Milestone is introducing Vision Language Model (VLM) as a Service, delivered via APIs. This initiative is aimed at allowing software developers, integrators, and Milestone partners to use the generative AI capabilities for other applications, irrespective of the underlying video management platform. The company sees this as an opportunity to enable a broader range of AI-powered solutions across different sectors.
"With this new XProtect plug-in, we are making advanced video intelligence accessible to cities, organizations, and operators everywhere who manage traffic systems - helping them unlock new levels of efficiency, safety, and insight. XProtect users will get access to state-of-the-art generative AI capabilities, and our partners will be able to build value on top of those new capabilities now available within XProtect. It truly marks a pivotal step in our mission to transform how the world manages and learns from visual data, responsibly and at scale," said Thomas Jensen, CEO, Milestone Systems.
