Summary
NVIDIA has outlined partner projects using AI agents, camera analysis, traffic simulation and digital twins for city and infrastructure services. The examples range from existing resident-service deployments to systems still being built or planned.
NVIDIA’s November 2, 2026 account describes partner projects applying AI to resident services, city camera feeds, traffic planning and industrial infrastructure. The examples use a mix of conversational agents, video-analysis tools and digital twins; they range from existing service deployments to systems still being built or planned.
AI services are being built on shared infrastructure
In Spain, 1MillionBot combines open-source ALIA models trained at Barcelona Supercomputing Center with NVIDIA Nemotron and other models for resident-service agents. NVIDIA says the agents support Madrid Digital, Administració Oberta de Catalunya and more than 150 deployments across Spain, handling tasks such as municipal procedures, appointment requests, urban-planning incidents and service inquiries.
In Poland, K2K is building shared AI infrastructure in Wrocław. The project uses NVIDIA’s Metropolis Blueprint for video search and summarisation to analyse existing camera streams for mobility, safety and event-response needs. The same infrastructure is intended to serve several city functions. In the San Francisco Bay Area, LiveX is developing conversational AI hosts for places including airports, hospitals and public events; NVIDIA says they can answer questions and take actions, with services tailored to local languages and needs.
Malaysia’s ITMAX is building another camera-based system. Its SamurAI V2 vision agents run on NVIDIA RTX PRO 6000 GPUs, with ITMAX targeting support for more than 17,000 cameras by 2027. The company also uses TeraSim, an open-source traffic-simulation platform from Laplace Intelligence, to test signal timing. NVIDIA reports peak-hour corridor travel-time reductions of up to 26% from this work.
An AI agent is a software system designed to respond to requests and carry out tasks, sometimes by using other tools or models. Here, the tasks range from answering service questions to searching video feeds. NVIDIA’s broader “physical AI” framing also includes systems that model or interact with real-world environments, such as traffic and industrial operations.
Digital twins bring simulation into planning and operations
At Dangote’s Lekki refinery in Nigeria, Honeywell Technologies’ Forge Plant Cognition uses Nemotron-powered agents to interpret operational data. The platform is built on Dell PowerEdge servers with NVIDIA RTX PRO accelerators. A digital twin of the plant and its processes is planned for simulation and training; NVIDIA says the current platform helps operators identify bottlenecks and anticipate equipment or process issues.
In Taiwan, Linker Vision and architecture firm MVRDV are using NVIDIA Omniverse and Nemotron to assess urban-planning scenarios with environmental, social and governance agents. NVIDIA says the system simulated 15-year outcomes for an urban forest canopy in 10 minutes, and reduced design iteration time from two weeks to less than an hour.
Other projects focus on building local capacity. At Morocco’s Tanger Med port, Cires Technologies is expanding a sovereign AI factory for planned uses including worker safety, congestion analysis and cargo routing. Barcelona Supercomputing Center’s AI Factory Accelerator Programme gives startups access to computing, technical support and training.
Together, the examples show a deployment approach built around shared computing resources and specialised AI applications, rather than separate infrastructure for every service. The projects are at different stages: some services are already deployed, while the refinery digital twin and ITMAX’s camera-support target are future plans.