Deploy Vision Models to NVIDIA Jetson Orin in Minutes | AI at the Edge

In this video, Roboflow Engineer Riaz Virani demonstrates how to provision, deploy, and manage computer vision applications on edge hardware using Roboflow Deployment Manager.

Starting with a fresh NVIDIA Jetson Orin J4012, Riaz shows how to go from raw hardware to a fully active computer vision stream in just a few minutes using a single install command.

= What you’ll learn =
Device Provisioning: How to add a new device and generate the installation command.
Remote Management: Monitoring device heartbeats, logs, and resource usage (GPU/CPU) from the cloud.
Stream Configuration: Setting up RTSP, USB, or Basler cameras and attaching a Roboflow Workflow for inference.
Live Visualization: Viewing the output of your computer vision model remotely to debug camera positioning or detections.
Fleet Scaling: How to replicate configurations across devices to manage deployments across multiple factories.

= Chapters =
00:00 - Intro: Running Vision AI on NVIDIA Jetson Orin
01:50 - Adding a Device in Deployment Manager
04:40 - Adding a Camera Stream & Workflow
08:00 - Device Logs & Alerts, Resource Monitoring, and Advanced Configurations

= Resources =
Documentation: https://docs.roboflow.com/deploy/devi...

Register for a future session: roboflow.com/webinar