NVIDIA is expanding AI services in cities worldwide through shared computing infrastructure, connecting municipal service agents, camera-based safety systems and digital twins. In a November 2, 2026 blog post, NVIDIA outlined projects ranging from resident inquiries in Spain to a Malaysian traffic-management deployment targeting more than 17,000 cameras by 2027 .
Key takeaways
NVIDIA’s city projects share computing resources across departments and municipalities.
Spain’s 1MillionBot combines ALIA and Nemotron models across more than 150 deployments.
NVIDIA reports peak-hour corridor travel-time reductions of up to 26% using TeraSim.
Taiwan’s urban-planning agents simulate 15-year forest-canopy outcomes in 10 minutes.
NVIDIA expands AI in cities through shared infrastructure
NVIDIA ’s approach gives city departments, neighboring municipalities and local developers access to a common computing foundation. The company says this lets them extend services without every organization acquiring and managing its own resources.
The expansion builds on initiatives presented at GTC Berlin. It combines physical AI and agentic AI applications for public services, infrastructure planning, maintenance and operational decisions, while keeping local governments in control of sensitive-data processing and use.
Municipal agents and conversational hosts
Spain’s 1MillionBot uses AI agents to handle municipal procedures, appointment requests, urban-planning incidents and service inquiries. Open-source ALIA models, developed through training at the Barcelona Supercomputing Center, are integrated alongside NVIDIA Nemotron and additional models within this system.
The deployments include Madrid Digital (Comunidad de Madrid) and Administració Oberta de Catalunya, within a network of more than 150 deployments across Spain , NVIDIA said.
In the San Francisco Bay Area, LiveX is developing interactive conversational hosts for airports, hospitals, local events and other public spaces. They can answer questions, understand service needs and take action, with adaptations for local languages and services.
LiveX uses NVIDIA Cosmos world models, Nemotron open models, Omniverse and Metropolis libraries, and RTX PRO GPUs.
Camera-based AI for safety and traffic management
Wrocław and Malaysian cities are using vision AI to turn camera feeds into information for mobility, safety and operational responses. These projects extend NVIDIA’s AI work in cities beyond resident-facing services.
In Wrocław, Poland, K2K is building shared infrastructure around NVIDIA accelerated computing, Cosmos world reasoning models and the NVIDIA Metropolis Blueprint for video search and summarization. The same infrastructure supports multiple city functions, including mobility, safety and event response.
Malaysia’s ITMAX packages infrastructure and agentic applications into long-term managed services. Its SamurAI V2 vision agents, developed with reference to Cosmos, Nemotron and the Metropolis blueprint, are scaling on NVIDIA RTX PRO 6000 GPUs toward the 2027 camera target.
ITMAX also integrates Laplace Intelligence ’s open-source traffic simulator, TeraSim, to test signal timing. NVIDIA reported reductions of up to 26% in peak-hour corridor travel times.
Digital twins for refinery operations and urban planning
Digital twins support infrastructure simulation and operational planning, with projects in Nigeria and Taiwan applying NVIDIA tools to plant performance and environmental assessment.
Nemotron-powered agents, deployed through Honeywell Technologies’ Forge Plant Cognition, are used to interpret operational data at Dangote’s Lekki refinery in Nigeria. The platform runs on Dell PowerEdge servers with NVIDIA RTX PRO accelerators, helping operators locate bottlenecks, anticipate equipment and process problems, and optimize performance.
An Omniverse-based digital twin of the refinery and its processes is planned for simulation and training .
In Taiwan, Linker Vision and Dutch architecture firm MVRDV are building environmental, social and governance agents with Omniverse and Nemotron. NVIDIA says they simulate 15-year urban forest-canopy outcomes in 10 minutes , shortening design iterations from two weeks to less than an hour.
Local AI capacity and the Barcelona showcase
Morocco’s Tanger Med port and the Barcelona Supercomputing Center are extending local AI capacity through sovereign computing and developer support.
At Tanger Med, Cires Technologies is expanding a sovereign AI factory for planned worker-safety, congestion-analysis, cargo-routing and port-operation applications. It is also assembling application providers to develop local models and tools for ports and cities.
The Barcelona Supercomputing Center’s AI Factory Accelerator Programme provides startups with computing access, technical assistance and training. BSC AI Factory supports research spanning agriculture, medical imaging, knowledge access and speech services.
NVIDIA will present these initiatives at the Smart City Expo World Congress in Barcelona, running November 3–5, 2026. Paolo Guglielmini, NVIDIA’s vice president of EMEA, is scheduled to give a keynote, followed by a panel moderated by Charbel Aoun, its director of smart city and spaces for EMEA.
Article produced with the assistance of artificial intelligence and reviewed by the editorial team.
