TracQi TrueNorth

OVERVIEW

Kotlin sensor fusion library that hits industry-standard positioning accuracy using nothing but consumer mobile phone sensors. Benchmarked across 90 sensor profile combinations and 6 track types, the 10-state filter achieves 5.0m RMSE on circuit tracks with standard 1 Hz phone GPS — matching published GNSS+MEMS IMU baselines — and 4.7m in urban canyon scenarios, beating the ~8m industry figure. At 10 Hz GPS update rates the error drops to 2.2m with standard GPS and 0.48m with RTK. A deterministic simulation harness with 100 Hz physics, configurable sensor noise profiles, and multi-rate measurement fusion provides repeatable benchmarking across GPS accuracy classes from standard through RTK Fixed. Ships with a KMP Compose calibration tuner and an Android reference app for real-device validation.