Evidence so far

An IMU (inertial measurement unit) measures turning and acceleration. The project's empirical evidence comes primarily from NASA survey-grade airborne IMUs. Stationary consumer crowdsourcing is a separate developing experiment.

NASA archive

All 53 selected high-speed stretches have been analyzed. Twelve have complete comparisons favoring globe over the specified flat disc under both processing assumptions; all twelve then favor the background-field globe over the zero-field globe. None has a complete flat preference. Forty-one remain unresolved because required numerical fits did not finish. These are model-fit preferences, without independently calibrated statistical significance.

Saved fits call those globe candidates “rotating” and “still.” Those names specify the background angular-rate term present in each prediction. The comparison does not identify what produces the measured rotation or establish a cosmology. A freely estimated field and a solar-versus-sidereal precision test remain follow-on work.

The source-aware findings give durations, speeds, instrument record types and partial results. The analysis roadmap tracks smaller numerical, timing, calibration and prediction checks. Overlapping streams and repeated flights are not counted as independent instruments.

Stationary consumer experiment

The Android app now guides still A/B reversals on a solid surface. Its preliminary analyzer estimates a signed horizontal reversal signal without latitude or an expected rotation amplitude. No real consumer unit has yet completed the required slow-rate validation here. Software tests establish specific behavior; they are not detections of a geographic field.

See the stationary method, instrument checks and validation plan. Geographic pooling and calibrated uncertainty remain future work.

Historical passenger-flight work

Two 1,000-attempt synthetic pilots and later route/turn studies remain recorded in the technical evidence and technical validation history. They explored aircraft collection and exposed motion, wind and sensor limitations. They do not validate the new ground protocol. No simulation campaign was rerun for this change.