Telemetry

Publishing sensor readings from a device, and viewing or querying them back out.

1. Publishing a reading

Publish one sensor at a time:

device.publish('temperature', 21.5)

Or batch multiple sensors into a single publish:

device.publish({ temperature: 21.5, humidity: 40 })

publish returns true if the reading was handed off, or false if the device is currently offline. Telemetry is fire-and-forget — a reading published while offline is dropped, not queued for later delivery. If you need to know whether a specific reading landed, publish again once the device reconnects.

2. Viewing a sensor’s readings

A telemetry sensor’s page shows its latest reading — the most recent value, updating automatically — separate from a historical chart you can switch between 1 hour, 6 hours, 24 hours, and 7 days. A sensor that hasn’t reported yet shows “No data yet.”

3. Querying across sensors

For anything beyond a single sensor, a project’s Analytics page lets you build ad-hoc queries:

• Pick one or more devices and sensors — an entire device’s sensors at once, or specific ones.
• Choose a time range, from 1 hour up to 30 days.
• Choose how to aggregate: Raw readings, Average, Minimum, Maximum, Sum, or Count.
• Optionally group results by Device, Sensor, and/or Time bucket (Minute/Hour/Day).
• View results as a Chart (Line/Area/Bar) or a Table, and export to CSV.

You always need to pick at least one device or sensor before running a query — there’s no “everything in this project” default.

Query limits: up to 100 device/sensor pairs per query. Raw-reading queries return at most 1,000 rows; grouped/aggregated queries return at most 5,000 rows. If your query’s true result set is larger, you’ll get back a capped set of rows rather than an error — narrow your time range or add grouping if you need to see the full picture.

Next steps

• Automations — reacting to the telemetry you publish.
• Device Provisioning — sensor types and the MQTT identity a device uses to publish.