--- name: lumeo-setup description: Install and configure Lumeo no-code video analytics gateway on NVIDIA Jetson. Covers account creation, gateway deployment, camera linking, and building a people detection pipeline. Requires Jetson with JetPack 5.1+ and internet access. --- # Getting Started with Lumeo on NVIDIA Jetson --- ## Execution model Run one phase at a time. After each phase: - Relay all command output to the user. - If output contains `[STOP]` → stop immediately, consult the failure decision tree below. - If output ends with `[OK]` → tell the user "Phase N complete" and proceed to the next phase. --- ## Prerequisites | Requirement | Minimum | |-------------|---------| | Hardware | NVIDIA Jetson device (tested on reComputer J4012) | | JetPack | 5.1+ with all SDK components installed | | Internet | Required on both Jetson and host PC | | Host PC | Windows, Linux, or Mac with browser access | | Lumeo account | Free account at https://console.lumeo.com/register | --- ## Phase 1 — Preflight Verify JetPack version and network connectivity. ```bash cat /etc/nv_tegra_release ping -c 3 lumeo.com ``` Expected: L4T version shown, ping succeeds. `[OK]` when both pass. `[STOP]` if no network. --- ## Phase 2 — Create Lumeo account (manual) 1. Visit https://console.lumeo.com/register — sign up with email and password. 2. Sign in at https://console.lumeo.com/login. 3. Enter an organization name and workspace name, click "Start using Lumeo". 4. Delete the default cloud gateway under Gateways (free accounts allow only one gateway). `[OK]` when Lumeo console is accessible and default cloud gateway is deleted. --- ## Phase 3 — Install Lumeo Gateway on Jetson Run the installer script on the Jetson device: ```bash bash <(wget -qO- https://link.lumeo.com/setup) ``` Respond to prompts (defaults are fine). When the installer completes successfully, type `install` at the prompt to deploy a new gateway container: ``` Enter the command: install ``` Enter a container name and log in with your Lumeo credentials when prompted. Verify the container is running: ``` Enter the command: list ``` Type `exit` to quit the script. ```bash # Verify gateway appears in Lumeo console under Gateways ``` `[OK]` when the gateway shows as online in the Lumeo console. `[STOP]` if installer fails or gateway does not appear. --- ## Phase 4 — Add cameras to the gateway 1. Connect a USB camera to the Jetson or ensure an ONVIF camera is on the same network. 2. In Lumeo console, click "Add Camera" under your gateway. 3. Click "Link" next to a discovered camera (or "Manually add cameras" if auto-discovery fails). 4. Enter camera name, provide credentials if needed, click "Connect camera". 5. Click the linked camera to verify a preview snapshot appears. For RTSP/HTTP streams: navigate to Deploy > Streams, click "Add input stream". `[OK]` when camera preview is visible in the Lumeo console. --- ## Phase 5 — Build a people detection pipeline 1. In Lumeo console, go to Design > Solutions, select "Basics - Detect Objects". 2. Keep default blocks (people detection model, tracker, WebRTC encoder) and click "Deploy". 3. Select your Jetson gateway, choose the configured camera, click "Deploy". 4. Wait for a green "running" status indicator. 5. Click the play button to view the output stream via WebRTC. `[OK]` when the live detection stream is visible with bounding boxes. --- ## Failure decision tree | Symptom | Action | |---------|--------| | Installer script fails to download | Check internet: `ping -c 3 link.lumeo.com`. Verify `wget` is installed: `sudo apt install wget`. | | Gateway not appearing in console | Ensure the default cloud gateway was deleted first (free tier limit). Re-run installer if needed. | | Camera not discovered | Click "Discover" again. For USB cameras, check `ls /dev/video*`. For ONVIF, verify same network/subnet. | | Pipeline deployment fails | Check gateway is online (green status). Verify camera is linked and producing preview snapshots. | | WebRTC stream not loading | Try a different browser (Chrome recommended). Check Jetson network firewall rules. | --- ## Reference files - `references/source.body.md` — full original Seeed tutorial with screenshots, camera setup details, and pipeline configuration (reference only)