Data Filtering and Traffic Reduction

This recipe demonstrates an in-vehicle IoT monitoring system connected to ThingsBoard Edge. The device sends readings from 10 sensors — distance, gas consumption, vehicle speed, engine temperature, ambient temperature, tire temperature, and tire pressure (four tires) — to the Edge. The Edge root rule chain filters that data and forwards only the distance readings to the cloud, reducing bandwidth while keeping full sensor history available locally.

• Completed the Getting Started guide.
• Familiar with the Edge Rule Chain Templates.
• ThingsBoard server is running and ThingsBoard Edge is connected to it.

Create a device named “In-vehicle monitoring system” on the Edge instance.

1. Log in to your ThingsBoard Edge instance and go to the Devices page.
2. Click ”+” in the top-right corner, then click Add new device.
3. Enter the name “In-vehicle monitoring system” and click Add.
4. The device now appears in the devices table.

Log in to your ThingsBoard Edge instance and navigate to the "Devices" page. Click on the "+" icon in the top right corner, then click "Add new device". Enter the name "In-vehicle monitoring system" and click "Add". The device "In-vehicle monitoring system" now appears in the devices table.

Once the device is created on Edge, verify it has been synchronized to the ThingsBoard server.

Log in to your ThingsBoard server instance and navigate to the "Devices" page. Verify that "In-vehicle monitoring system" is listed in the devices table.

Update the Edge Root Rule Chain on the ThingsBoard server to add a transformation script that filters incoming messages and forwards only the distance reading to the cloud.

Here is the final rule chain configuration you are aiming for:

The transformation script creates a new message containing only the distance field:

var newMsg = {};
newMsg.distance = msg.distance;
return { msg: newMsg, metadata: metadata, msgType: msgType };

1. Log in to your ThingsBoard server. Go to Edge management → Rule chain templates and open Edge Root Rule Chain.
2. Search for nodes containing “script”. Drag a Script (Transformation) node onto the canvas.
3. Enter the node name “Transform Incoming Message” and paste the JavaScript snippet above.
4. Delete the connection between “save timeseries” and “push to cloud”: select the connection, then click X.
5. Drag a new connection from “save timeseries” to “Transform Incoming Message”. Choose Success and click Add.
6. Drag a connection from “Transform Incoming Message” to “push to cloud”. Choose Success and click Add.
7. Click Apply Changes to save.

Navigate to Edge management → Rule chain templates and open "Edge Root Rule Chain". Search for nodes containing "script". Drag the Script node (Transformation) to the rule chain. Enter the node name "Transform Incoming Message" and paste the JavaScript code. Delete the connection between "save timeseries" and "push to cloud" by selecting it and clicking X. Drag a connection from "save timeseries" to the transformation script node. Choose "Success" from the dropdown list and click "Add". Drag a connection from "Transform Incoming Message" to the "push to cloud" node. Choose "Success" from the dropdown list and click "Add". Click "Apply Changes" to save the updated rule chain.

Now verify that the updated rule chain has been synchronized to the Edge instance.

Log in to your ThingsBoard Edge instance and navigate to the "Rule chains" page. Open "Edge Root Rule Chain" to verify the changes. The rule chain configuration on Edge matches what was configured on the server.

Get the device access token from Edge, then run the telemetry generator script.

1. In the ThingsBoard Edge UI, go to Entities → Devices.
2. Click on “In-vehicle monitoring system” to open device details.
3. Click Copy access token — the token is copied to your clipboard.

Open the Devices page in the ThingsBoard Edge instance. Click on the "In-vehicle monitoring system" row to open device details. Click "Copy access token". The token is copied to your clipboard.

Install the MQTT Python library:

pip install paho-mqtt

Download the telemetry generator script, update the three placeholders at the top, then run it:

import paho.mqtt.client as mqtt
import json
import random
import time

ACCESS_TOKEN = YOUR_ACCESS_TOKEN # REPLACE with your device access token, e.g. 'edge_vehicle'
THINGSBOARD_EDGE_HOST = YOUR_TB_EDGE_HOST # REPLACE with your ThingsBoard Edge host, e.g. 'localhost'
THINGSBOARD_EDGE_MQTT_PORT = YOUR_TB_EDGE_MQTT_PORT # REPLACE with your Edge MQTT port, e.g. 11883 or 1883
DISTANCE = 0

def on_connect(client, userdata, rc, *extra_params):
    client.subscribe('v1/devices/me/rpc/request/+')
    client.publish('v1/devices/me/telemetry', get_telemetry(), 1)
    print(get_telemetry())

def get_telemetry():
    return json.dumps(device)

while(True):
    device = {
        'distance': DISTANCE,
        'gas_consumption': random.randint(10, 15),
        'vehicle_speed': random.randint(60, 70),
        'engine_temperature': random.randint(195, 220),
        'ambient_temperature': random.randint(65, 80),
        'tire_temperature': random.randint(90, 110),
        'tire_pressure_front_left': random.randint(30, 35),
        'tire_pressure_front_right': random.randint(30, 35),
        'tire_pressure_rear_left': random.randint(30, 35),
        'tire_pressure_rear_right': random.randint(30, 35),
    }
    DISTANCE = DISTANCE + random.randint(0, 100)

    client = mqtt.Client()
    client.on_connect = on_connect
    client.username_pw_set(ACCESS_TOKEN)
    client.connect(THINGSBOARD_EDGE_HOST, THINGSBOARD_EDGE_MQTT_PORT, 60)
    try:
        time.sleep(5)
        client.loop()
    except KeyboardInterrupt:
        client.loop_stop()
        print("Exited!")

python mqtt-generator.py

Verify that the device is receiving all 10 sensor readings on Edge:

Navigate to the "Devices" page on the Edge instance. Click on "In-vehicle monitoring system" to open device details. All 10 sensor readings are visible in the Latest telemetry tab on Edge.

Verify that only distance readings are pushed to the ThingsBoard server:

Log in to the ThingsBoard server and open the "Devices" page. Click on "In-vehicle monitoring system" to view its details. The "Latest telemetry" tab on the server shows only distance readings.

Create a dashboard on the ThingsBoard server to visualize the total distance traveled.

1. Log in to the ThingsBoard server. Navigate to Dashboards.
2. Click ”+” and select Create new dashboard. Enter the title “Edge Vehicle” and click Add.
3. Open the “Edge Vehicle” dashboard to begin editing.
4. Click the Entity aliases icon in the toolbar. Click Add alias.
5. Set Alias name to edge device, Filter type to Single entity, Type to Device, and select “In-vehicle monitoring system”. Click Save.
6. Click Add Widget, find Digital gauges in the Widgets Bundle, and select the gauge widget.
7. Set Type to Entity, Entity Alias to edge device, and Data Key to distance. Click Add.
8. Open the widget edit panel. On the Appearance tab, set Max value to 10000 and Unit title to MLS. Click Add.

Log in to the ThingsBoard server and navigate to the "Dashboards" page. Click "+" and select "Create new dashboard". Enter the title "Edge Vehicle" and click "Add". Click on the "Edge Vehicle" dashboard to open it. Click the "Entity aliases" icon in the toolbar. Click "Add alias". Set Alias name to "edge device", Filter type to "Single entity", Type to "Device", and select "In-vehicle monitoring system". Click "Add Widget". Find "Digital gauges" in the Widgets Bundle and click to open the widget list. Select the gauge widget from the list. Set Type to "Entity", Entity Alias to "edge device", and Data Key to "distance". Click "Add". Click the edit icon in the upper right corner of the widget. Select the "Appearance" tab. Set Max value to "10000". Enter "MLS" as the unit title and click "Add". Dashboard widget showing the distance reading from the in-vehicle monitoring system.

You have successfully sent telemetry from the device, saved all sensor readings on the Edge, pushed only the filtered distance data to the cloud, and visualized it on the dashboard.