Urban centers generate vast amounts of data through cameras, sensors, and operational systems, but effectively converting this data into meaningful action remains a significant challenge. NVIDIA is addressing this with physical AI—digital twins and simulation models—and agentic AI, which deploys interactive AI agents to deliver public services and insights directly to city departments and residents.
Agentic AI applications are increasingly prevalent across cities such as Madrid, Barcelona, and San Francisco Bay Area, where AI-driven bots and vision agents assist with municipal procedures, service inquiries, and safety monitoring. Notably, Madrid Digital utilizes open-source ALIA models combined with NVIDIA Nemotron to handle over 150 different municipal service deployments, enabling responsive and tailored citizen engagement.
Vision AI agents harness existing city camera streams to provide transportation and safety insights. In Wrocław, Poland, AI infrastructure powered by NVIDIA's Cosmos and Metropolis Blueprint technologies enables multi-departmental access to actionable intelligence derived from video data, enhancing mobility management and emergency coordination.
Digital twins represent another core technological advance. These real-time, simulation-based replicas of city infrastructures facilitate rapid planning and operational decisions with reduced disruption. For instance, Honeywell Technologies is implementing a digital twin of Nigeria’s Dangote Lekki refinery, coupled with AI agents analyzing operational data to optimize plant performance and anticipate maintenance needs.
Cities are building shared AI computing infrastructure, allowing multiple departments and neighboring municipalities to leverage common resources without duplicative investment. In Malaysia, ITMAX integrates agentic AI with digital twins and open-source traffic simulations to manage over 17,000 cameras, successfully reducing peak-hour corridor travel times by up to 26%.
While these AI systems demonstrate significant benefits in speed, efficiency, and scalability, the complexity of managing sensitive urban data demands strict governance. NVIDIA emphasizes maintaining local government control over data processing, ensuring privacy and sovereignty are preserved even as AI adoption expands.
Additionally, ecosystem support programs like the Barcelona Supercomputing Center's AI Factory Accelerator help nurture local AI startups, fostering innovation across sectors beyond city services, including agriculture and medical imaging. This broader AI ecosystem enhances the potential reach and impact of physical and agentic AI technologies for urban environments.
Source & further reading
Cities Worldwide Build and Scale AI Services With NVIDIA Physical and Agentic AI ↗Source published 2026-11-03. This report was prepared with AI assistance from the linked primary-source material. Claims and limitations should be checked against the original announcement or research.