n8n node

n8n is a workflow automation platform that lets you connect ThingsBoard to hundreds of other services.

The ThingsBoard n8n Node gives you direct access to your IoT entities, telemetry, and alarms from n8n workflows:

• Manage IoT devices, assets, and customers directly from n8n workflows.
• Access and manipulate telemetry data in real-time with attribute and time-series operations.
• Monitor alarms and create automated responses based on alarm severity and type.
• Navigate entity relationships to understand your IoT infrastructure topology.
• Build AI-powered IoT automation using n8n AI Agents with natural language commands.

Here’s what you can build in 5 minutes — a real-world data pipeline that exports IoT telemetry to cloud storage.

Scenario: Every day at midnight, automatically export the last 24 hours of device telemetry to AWS S3 as JSON files for data warehousing, analytics, or compliance archival.

The complete workflow: Schedule trigger → Code → Get device → Get timeseries → Convert to file → AWS S3 upload. Successful execution: telemetry data exported to S3 as a JSON file.

This is a production-ready pattern for IoT data archival, analytics pipelines, and compliance logging. See Example 2 for the full step-by-step walkthrough.

ThingsBoard’s Rule Engine handles IoT-specific automation. When you need to connect to external systems that ThingsBoard doesn’t integrate with natively, n8n fills the gap.

• ThingsBoard instance — self-hosted or ThingsBoard Cloud.
• Authentication credentials — API key or username and password (deprecated) with appropriate permissions. See API Keys for details on generating API keys.
• n8n instance — local installation, Docker-based, or n8n Cloud (the ThingsBoard node is verified and available).

Make sure n8n is already running. See the official n8n docs if you need to install it first.

1. Open n8n in your browser.
2. Click the three dots (bottom-left) → Settings → Community Nodes.
3. Click Install a community node.
4. Enter: @thingsboard/n8n-nodes-thingsboard and click Install.
5. Refresh your browser.

Open n8n, click the three dots (bottom-left), and select Settings. Go to Community Nodes and click Install a community node. Enter @thingsboard/n8n-nodes-thingsboard and click Install. Wait for installation to complete, then refresh the browser.

Skip this if you are only using the node in regular workflows (not with AI agents).

To let AI agents use ThingsBoard as a tool, add this environment variable before starting n8n:

N8N_COMMUNITY_PACKAGES_ALLOW_TOOL_USAGE=true

• Docker: Add to your docker-compose.yml environment section.
• npm/local: N8N_COMMUNITY_PACKAGES_ALLOW_TOOL_USAGE=true n8n start

Connect n8n to your ThingsBoard instance. There are two authentication options:

1. Click + (top-left) → Credentials.
2. Search for ThingsBoard API Key → Continue.
3. Fill in the required fields:
   • Base URL: Your instance URL (e.g., https://thingsboard.cloud).
   • Connect using: API Key.
   • API Key: Generated from your ThingsBoard instance. See API Keys for details.
4. Click Save.

1. Click + (top-left) → Credentials.
2. Search for ThingsBoard Username/Password → Continue.
3. Fill in the required fields:
   • Base URL: Your instance URL (e.g., https://thingsboard.cloud).
   • Connect using: Username/Password.
   • Username: Your email.
   • Password: Your password.
4. Click Save.

The credentials are encrypted and stored securely by n8n. You can reuse the same credentials across multiple ThingsBoard nodes in different workflows.

Click + and select Credentials from the dropdown. Search for ThingsBoard API to connect to and click Continue. Fill in the Base URL — your ThingsBoard instance URL (without a trailing slash), and the API Key — generated from your ThingsBoard instance. Click Save to store the credentials. The credentials are saved and the connection is tested successfully. You can reuse the same credentials across multiple ThingsBoard nodes in different workflows.

1. Open n8n in your browser.
2. Click Create Workflow (top-right).
3. Click the + button on the workflow canvas.
4. Type “ThingsBoard” in the search field.
5. Verify the ThingsBoard node appears with available actions.

Click Create Workflow in the top-right corner. Click the + button on the workflow canvas. Type ThingsBoard in the search field — the node should appear. A list of available ThingsBoard actions is displayed.

If you don’t see the node: verify N8N_COMMUNITY_PACKAGES_ENABLED=true is set, check that npm installation completed without errors, restart n8n, and check logs.

Drag the ThingsBoard node onto your canvas and configure the operation (get devices, save telemetry, create alarms, etc.).

Best for: scheduled tasks (daily reports), webhooks (external triggers), multi-step workflows (get device → fetch telemetry → send to Slack).

Give an AI agent access to ThingsBoard operations. Users ask questions in plain English instead of building workflows.

Best for: non-technical users querying entities/data, support teams troubleshooting.

Requires AI agent configuration.

Scenario: Ask questions in plain text to check entities, query data, or update configurations.

1. Create a new workflow with an AI Agent node. Choose your AI model (Gemini, Anthropic, ChatGPT, etc.) and configure memory settings.
2. Select ThingsBoard from the Tools list. Add Get devices and Get timeseries operations.
3. For tools with required fields, the AI model determines parameter values from conversation context automatically.
4. Open the chat interface and write your natural language query.

Create a workflow with AI Agent node and choose your AI model. Select ThingsBoard tools: Get devices, Get timeseries. The AI model intelligently determines parameter values from context. Open the chat and write your query in natural language.

How it works:

1. User asks: “Which freezers in the warehouse are showing temperatures above -10°C?”
2. AI Agent automatically calls Get devices to find freezer devices, then Get timeseries for temperature readings, filters results, and responds in plain language.
3. No code or API knowledge needed.

Example queries:

• “Show me all devices that haven’t sent data in the last 24 hours”
• “What’s the current temperature of Freezer 3?”
• “Which sensors are in Building A?”
• “Get me the last week of humidity data for all warehouse sensors”

Scenario: Every night, export device telemetry to S3 for data warehousing, enriched with metadata and ready for analytics tools.

Why use n8n instead of ThingsBoard’s native export:

• Multi-destination — same data to S3 + Snowflake + email report in one workflow.
• Data transformation — enrich telemetry with business context (customer names, locations from CRM).
• Custom formats — convert to Parquet, Avro, or CSV with specific schemas.
• Integration chains — after S3 upload, trigger AWS Lambda, update tracking database, send Slack notification.

Step 1: Add Schedule Trigger

Click Add first step → Schedule Trigger. Set trigger to run every day at 00:00 (midnight UTC).

Create a new workflow and select Schedule Trigger. Configure to run every day at 00:00 UTC.

Step 2: Add Code node for time range

ThingsBoard’s API requires timestamps in milliseconds. Add a Code node:

const DAY_MS = 24 * 60 * 60 * 1000;
const endTsMs = new Date($input.first().json.timestamp).getTime();
const startTsMs = endTsMs - DAY_MS;
return [{ startTsMs, endTsMs }];

Add a Code node after the Schedule Trigger. Paste the time calculation code and execute to test.

Step 3: Get device by name

Click + → ThingsBoard → Get a device by name. Enter your device name (e.g., “Temperature Sensor 1”).

Add ThingsBoard node and select Get a device by name. Enter the device name and execute to retrieve the device object.

Step 4: Get timeseries data

Click + → ThingsBoard → Get timeseries. Drag values from the INPUT panel:

• Entity ID and Entity Type: From “Get a device by name” output.
• Start/End timestamps: From the Code node output.
• Keys: Telemetry keys you want (e.g., “temperature,humidity”).

Add ThingsBoard Get timeseries node. Configure with device ID, entity type, and time range from previous nodes.

Step 5: Upload to AWS S3

Click + → AWS S3. Configure: Operation = Upload, Bucket Name, File Name with datestamp expression (e.g., telemetry-{{ $now.format('yyyy-MM-dd') }}.json), and AWS credentials.

Add AWS S3 node with upload configuration. Telemetry data successfully uploaded to S3.

Result: A complete automated pipeline that runs every night, fetches 24 hours of telemetry, and uploads to S3 with predictable naming (telemetry-2024-12-23.json).

Scenario: A customer portal has a “Get Device Status” button. When clicked, it fetches real-time telemetry for any device — no hardcoded IDs.

This pattern enables: customer portals (each user sees their own devices), chatbots (“check temperature in Room 405”), mobile apps (query any device by name).

Step 1: Set up the trigger

Use a manual trigger to simulate a webhook. Click Add first step → Trigger manually, then configure with test JSON:

{
  "deviceName": "Refrigerator"
}

Click Add first step and select Trigger manually. Click on the trigger node to open configuration. Click the pencil icon (top-right) to edit input. Paste the JSON and save changes.

Step 2: Find the device

Click + → ThingsBoard → Get device by name. Drag deviceName from the INPUT panel into the Name field — this creates a dynamic link.

Click + next to the trigger node. Find and select ThingsBoard. Choose Get a device by name. Drag deviceName from the INPUT panel into the Name field. Click Execute step to test. Returns the full device object including ID, type, and metadata.

Step 3: Discover telemetry keys

Click + → ThingsBoard → Get timeseries keys. Drag Entity ID and Entity Type from the INPUT panel.

Add ThingsBoard Get timeseries keys node. Entity ID and Entity Type are set dynamically. Execute to discover available keys.

Step 4: Fetch telemetry data

Click + → ThingsBoard → Get timeseries. Configure:

• Entity ID and Entity Type from Step 2.
• Keys from Step 3 — join as comma-separated: {{ $json.timeseriesKeys.join(',') }}.

Add ThingsBoard Get timeseries action. Configure with device ID, entity type, and keys. Execute to fetch data.

Result: A completely reusable workflow. Change the input from “Refrigerator” to “Freezer 3” and it works automatically — no hardcoded IDs, no changes needed.

Accepts a device name and finds the corresponding device. Discovers available telemetry keys. Retrieves telemetry values for the specified time window.

Next steps: Replace the manual trigger with a Webhook to connect your customer portal, add Slack or Email output, use a Schedule trigger for monitoring, or connect to an AI Agent for chat-based queries.

• npm package: @thingsboard/n8n-nodes-thingsboard
• GitHub: thingsboard/thingsboard-n8n-node
• n8n docs: docs.n8n.io
• ThingsBoard REST API: REST API Reference