At COMPUTEX 2025, Taiwan-based AI software leader Linker Vision revealed its latest breakthroughs in large-scale urban AI deployments. These innovations leverage deep NVIDIA technology integration to advance the Linker VisionAI Platform—unlocking new capabilities in digital twin simulation, agentic AI, and synthetic data generation (SDG) for smart city applications.
From Simulation to Deployment: The NVIDIA Three Computer Solution
Linker Vision employs NVIDIA’s “three computer solution” framework—simulate, train, deploy—to deliver end-to-end AI lifecycle support for complex urban operations. By utilizing NVIDIA Omniverse libraries, Linker creates high-fidelity digital twins that replicate dynamic city environments with precision. NVIDIA Cosmos world foundation models fuel synthetic data generation, accelerating the creation of diverse training scenarios that improve AI model robustness.
These tools enable seamless real-to-sim and sim-to-real workflows, enhancing AI responsiveness for applications including traffic monitoring, disaster response, and public safety.
AI-Powered Video Analytics at Scale
Leveraging NVIDIA AI Blueprint for video search and summarization (VSS) alongside NVIDIA Metropolis, Linker Vision trains and deploys AI agents capable of analyzing thousands of live camera streams. This facilitates real-time situational awareness, automated anomaly detection, and intelligent decision-making critical to smart city management.
Breaking Data Silos with Vision Language Models
A core component of Linker Vision’s platform is its vision language model (VLM), which translates complex video data into contextual insights. By unifying disparate sensor and camera inputs from multiple agencies, this AI system fosters cross-domain collaboration through a centralized processing pipeline.
City governments use VLMs to detect traffic violations, unsafe behaviors, and other anomalies in real time—generating alerts and optimizing incident response. Linker Vision is partnering with Kaohsiung City and Chunghwa Telecom to build scalable AI infrastructure supporting real-time governance and digital city operations.
Scaling Across Urban Domains and Enterprises
This AI infrastructure model is extending beyond municipal governments to state-owned enterprises and commercial sectors. The goal: to scale intelligent video analytics and AI-driven governance across a broad range of urban applications—transforming smart city operations with scalable, actionable intelligence.
Linker Vision’s showcase at COMPUTEX 2025 marks a pivotal step toward integrating AI deeply into urban infrastructure—empowering smarter, safer, and more efficient cities worldwide.
