# Testing and calibration

1. Confirm the wiring and power are disconnected while wiring.
2. Open the app, connect the CH340, and verify valid frames in Diagnostics.
3. Start with simulation mode to exercise the UI without hardware.
4. For roadside testing, set the counting-line distance to the physical line in front of the radar. Set plausible minimum and maximum speeds and choose the expected direction.
5. Compare experimental counts with a manual tally. Adjust mounting angle, line distance, minimum observations and track timeout.

The tracker requires several observations and a line crossing; it suppresses repeated counts but is not certified traffic measurement. Radar observations are not proof that an object is a car, and the protocol does not provide vehicle classification.

Synthetic simulation fixtures are generated in `js/ld2451-parser.js`. No physical fixture is bundled because raw captures can contain sensitive location/time information.