Unit 8: Sensors and Measurement

A sensor produces evidence. It does not produce truth.

Make the measurement chain visible

A Sensor Does Not Know the Truth separates physical reality from the value a device reports. Calibration, Noise, and Sampling examines why that value changes. Build a Measurement Chain traces one value from transduction through conversion and validity logic.

The essential question is not "does the sensor work?" It is:

Compared with what reference, over what range, at what rate, and with what uncertainty?

Preserve the ugly data

Build a calibration/noise record and a complete measurement-chain diagram.

Do not throw away the samples that disagree with your expectation. Outliers, drift, saturation, and dropouts are exactly where sensor assumptions become visible.

Useful language includes calibration, precision, accuracy, bias, noise, sample rate, resolution, validity, and stale data.

Live sensors are optional

Supplied datasets and simulated sensors can teach the full reasoning path. If you do use physical hardware, preserve the reference condition and measurement method so another person can understand what the numbers mean.