Sample data · CC0 / public domain

Sample iot sensor readings CSV.

A telemetry stream: timestamped readings from forty devices with temperature, humidity, battery level and status. The shape of every time-series ingestion and downsampling demo.

Download

Grab a file — or generate a big one.

The small files are static downloads. The large ones are generated in your browser from the same fixed seed, so every copy of sensors-100000.csv on earth is byte-identical — reproducible test data with no 60 MB download.

100 rows · 5 KB 1,000 rows · 44 KB

→ Open this dataset in the dashboard builder  ·  → Open in the CSV editor

Preview

First rows.

timestampdevice_idtemperature_chumidity_pctbattery_pctstatus
2025-01-01T00:00:02Zdev-1836.1624.98ok
2025-01-01T00:00:32Zdev-3431.3672.649ok
2025-01-01T00:01:05Zdev-1139.9484.615ok
2025-01-01T00:01:33Zdev-1922.5334.697ok
2025-01-01T00:02:07Zdev-1831.4368.948ok
2025-01-01T00:02:31Zdev-3330.9388.114ok
2025-01-01T00:03:03Zdev-1229.6072.774ok
2025-01-01T00:03:32Zdev-2532.8962.432warning
Schema

Columns.

columndescription
timestampReading time (UTC, ~30s cadence)
device_idReporting device
temperature_cSensor temperature (°C)
humidity_pctHumidity (%)
battery_pctBattery level
statusDevice status
About this dataset

What it models.

Timestamps advance monotonically (one reading per 30 seconds with jitter), so resampling, windowing and gap-detection demos behave like production telemetry.

A small share of readings are warning/error status with low battery — enough signal for alerting-logic demos.

Good for: Downsampling / moving-average demos · Time-window aggregation tests · Alert-threshold examples.

License: CC0 / public domain — use it anywhere, no attribution needed.

Common questions
  • ·

    Are timestamps strictly increasing?

    They advance ~30 seconds per row with a few seconds of jitter — monotonic enough for windowing demos, jittered enough to be realistic.

  • ·

    What license is this under?

    CC0 (public domain). Use it in tutorials, tests, courses, screenshots and products — no attribution required.

  • ·

    Is the data deterministic?

    Yes — every size is generated from a fixed seed, so the same file is byte-identical for everyone, forever. Reproducible tests, stable teaching materials.

More sample data