Sensors, measurements, units, and calibration
Learn what a sensor contributes to a measurement—and what metadata you still need before comparing values.
Why this matters
A number without units, method, or sensor context can be useless or misleading. Two sensors measuring the same phenomenon may also have different ranges, calibration histories, accuracy, or sampling behavior.
Learn it
A sensor is the device or component that responds to a phenomenon. A measurement is a value produced through an observation process. Units tell you the measurement scale. Calibration relates instrument response to known references; it does not mean every future reading is perfect. Sampling rate tells you how frequently measurements are taken or reported.
Work through an example
18.4 means little by itself. '18.4 °C from sensor S-12 at 17:00 UTC, calibration checked on date X' is far more useful.
sensor_id: S-12
value: 18.4
units: °C
observed_at: 2026-09-30T17:00:00ZPractice
- 1Take a measurement you already collect and list value, units, sensor, method, and time.
- 2Identify which metadata comes from the sensor and which belongs to each observation.
- 3Decide what information is needed before comparing two sensors.
What to watch for
Do not silently convert units without recording the transformation. Preserve source values when provenance or reproducibility matters.
Where this takes you
This is foundational for weather, water, vehicle, environmental, industrial, and classroom observations.