Step #5: Deploy a Sample Application

Deploy Vision AI at the Edge of the Network (Latest Version)

Now that the local installer has helped get the system paired to Fleet Command, it is now online at the intersection and ready to start running applications. As the Fleet Command administrator, you’ll deploy a simple AI application to perform traffic monitoring, analyzing video feeds in real-time, then identifying and counting the cars at this intersection.

Applications can be hosted on your NGC Private Registry, or directly on the NGC Catalog. This allows you to customize your own application to suit your specific needs, and create your own models, helm charts, etc. to use with your application.

The sample application is already uploaded to NGC, the next step is to import it into Fleet Command.

  1. On the Fleet Command Application page, click on Add Application.

    fc-012.png

  2. Enter the information from the Helm Chart URL details below into the appropriate fields in the Add Application screen.

    • Display Name: Car Counter

    • Description: Evaluate traffic patterns at intersections

    • Helm Chart Name: video-analytics-demo

    • Helm Chart Version: 0.1.8

    • Helm Repo URL: https://helm.ngc.nvidia.com/nvidia

    Important

    There is no trailing slash in the Helm Repo URL.

    fc-013.png

    fc-014.png

    fc-015.png


  3. Next, you go to the Deployment page and click on Create Deployment.

    fc-016.png

  4. Enter in a name for the deployment, select the location and application you created, check the box at the bottom acknowledging the deployment, then click Deploy to start the application deployment.

    fc-017.png

  5. The application deployment should then complete within a few minutes.

    fc-018.png

You have just deployed an AI application to the Edge! In the next step we will verify the application is available using the Fleet Command console.

© Copyright 2022-2023, NVIDIA. Last updated on Jan 10, 2023.