Cities Worldwide Harness NVIDIA Physical and Agentic AI to Enhance Urban Services and Planning

Image: NVIDIA · Source

NVIDIA's physical and agentic AI technologies enable cities globally to transform data from cameras and sensors into actionable insights, improving public services and infrastructure planning while maintaining data control.

Cities around the world generate vast amounts of data through cameras, sensors, and operational systems, but converting this information into actionable insights for urban planning, maintenance, and public services remains a significant challenge. NVIDIA is addressing this gap by promoting the adoption of physical and agentic AI technologies, which offer innovative ways to turn data into practical applications while allowing local governments to retain control over sensitive information.

Building on initiatives introduced at GTC Berlin, NVIDIA and its partners are expanding their AI solutions to cities internationally, leveraging shared computing infrastructure that enables city departments, neighboring municipalities, and local developers to collaborate without each needing to invest heavily in resources individually.

Agentic AI is being used extensively to enhance public services. For example, in Spain, the 1MillionBot system combines open-source ALIA models trained at the Barcelona Supercomputing Center with NVIDIA Nemotron and other models to power resident-service agents managing municipal procedures and inquiries across over 150 deployments nationwide, including in Madrid and Catalonia.

In the San Francisco Bay Area, LiveX utilizes NVIDIA Cosmos world models, Nemotron open models, and Omniverse and Metropolis libraries on NVIDIA RTX PRO GPUs to develop conversational AI hosts that interact with people in airports, hospitals, and local events, providing tailored responses and services.

Vision AI agents are also transforming city camera feed data into operational insights for mobility, safety, and emergency coordination. For instance, in Wrocław, Poland, K2K employs NVIDIA accelerated computing and world reasoning models to convert video surveillance into actionable intelligence for multiple city functions.

Malaysia's ITMAX deploys scalable agentic AI applications like SamurAI V2 vision agents on NVIDIA RTX PRO 6000 GPUs to manage over 17,000 cameras by 2027, contributing to traffic management improvements such as up to a 26% reduction in peak-hour travel times through integrated simulation platforms.

Beyond street-level services, digital twins based on NVIDIA Omniverse facilitate infrastructure planning and operational simulation. Honeywell Technologies uses AI agents and digital twin technology powered by Nemotron and Omniverse at the Dangote Lekki refinery in Nigeria to optimize plant operations and anticipate issues.

Similarly, in Taiwan, Linker Vision and architecture firm MVRDV utilize these tools to simulate urban environmental outcomes rapidly, accelerating urban forest canopy design iterations from weeks to under an hour.

NVIDIA is also fostering broader local AI ecosystems. At Morocco's Tanger Med port, Cires Technologies is building a sovereign AI factory aimed at enhancing worker safety, congestion management, and port operations, while supporting local application development. The Barcelona Supercomputing Center’s AI Factory Accelerator Program provides startups access to sovereign computing resources and technical support, aiding innovation in sectors like agriculture and healthcare.

These initiatives underscore NVIDIA’s commitment to empowering cities with advanced physical and agentic AI to improve urban management and services. Detailed insights into these projects and NVIDIA’s vision will be featured at the upcoming Smart City Expo World Congress in Barcelona, where NVIDIA’s Vice President of EMEA and other experts will present keynotes and panels on AI-driven urban transformation.