Applications

A city that notices.

Surveillance cameras on every corner and sensors on every street generate terabytes an hour. YUAN's SmartNVR, SmartVMS, and SmartFlow turn that deluge into decisions — at the edge, in real time.

275: TOPS of edge compute on VPP6N0-MB AGX
3: SmartApps in one ecosystem — SmartNVR, SmartVMS, SmartFlow
4K+: surveillance resolutions processed at the edge

The challenge

The data arrives faster than anyone can watch it.

City managers face a deluge: surveillance cameras in every corner and traffic sensors linking each street, continuously generating TB-levels of valuable information. Nobody is watching most of it, and by the time footage is reviewed, the moment it recorded has passed.

As surveillance moves from 4K towards even higher resolutions, the demands on data processing hardware become unprecedented. Hauling every frame to the cloud is too slow and too expensive; the intelligence has to move to where the cameras are.

The data arrives faster than anyone can watch it.

The architecture

SmartNVR at the edge, SmartVMS in the cloud.

YUAN's SmartNVR VPP6N0 AGX processes data on-site, streamlining traffic and public safety operations without over-reliance on the cloud — rapid, decisive action where urban efficiency is won or lost. In sync, SmartVMS streams to the cloud, where city managers get centralised dashboards, live monitoring, map integration, and unified management across multiple sites.

SmartFlow completes the ecosystem with people analytics. Together the three form a complete edge-to-cloud AI platform, powered at its core by the NexVDO SDK — real-time insights, anomaly detection, and system-wide control.

SmartNVR at the edge, SmartVMS in the cloud.

The watchpoints

Four jobs the AI never gets tired of.

Pedestrian and vehicle flow: the models identify vehicle types and analyse large-scale movement in real time, so the city can adjust activity and allocate resources intelligently. Public safety: abnormal behaviour — unauthorised entry, a suspicious abandoned package — is identified within milliseconds and reported straight to law enforcement.

Traffic enforcement: violations are identified and analysed automatically, raising the degree of automation in city management. Disaster prevention: scene recognition and object detection spot the signs of flood or fire and locate the problem area before it spreads.

In the field

From Japanese mountainsides to industrial CCTV.

COMJO's AIJO Site Monitor runs on YUAN's Pandora AI platform with NVIDIA Jetson Orin NX Super Mode, delivering real-time wildlife detection and alerts — including bear detection — now in field deployment with Japanese government agencies. Swappable AI modules extend the same box to helmet compliance and meter reading.

Fogsphere adopts YUAN's EDG6N0-S T5X, powered by the NVIDIA Jetson Thor T5000 module, for real-time safety monitoring, intelligent CCTV analytics, and PPE detection across industrial and urban environments. Same platform, different streets.