Data set | Ambiguity Resolution | Overall positioning performance (fix & float epochs) | |||||||
---|---|---|---|---|---|---|---|---|---|
3D | Horizontal | Vertical | |||||||
PV (%) | Pf (%) | d95 (cm) | RMSE (cm) | d95h (cm) | RMSE (cm) | d95v (cm) | RMSE (cm) | ||
May 9 | Unaided | 77.31% | 0.33% | 742.1 | 1799.6 | 471.4 | 1713.3 | 209.3 | 550.5 |
Consumer-grade IMU | 94.56% | 0.19% | 20.2 | 34.4 | 16.4 | 28.6 | 7.3 | 19.2 | |
Industrial-grade IMU | 97.28% | 0.21% | 13.8 | 20.6 | 10.6 | 18.9 | 4.0 | 8.1 | |
May 12 | Unaided | 76.94% | 0.49% | 860.5 | 571.9 | 525.5 | 406.1 | 467.9 | 402.7 |
Consumer-grade IMU | 98.63% | 0.56% | 13.7 | 7.7 | 10.6 | 6.0 | 5.7 | 4.8 | |
Industrial-grade IMU | 97.74% | 0.43% | 12.3 | 13.8 | 9.7 | 12.5 | 4.9 | 5.8 |
Note: Position errors are reported relative to the TEX-CUP ground truth trajectory, which was generated with a survey-grade mobile mapping system as described in Section 5.1, with a sel-freported accuracy varying between 2 and 15 cm (1σ) along the route. Unaided indicates the use of the motion model of Humphreys et al. (2020b) in lieu of inertial tight coupling, as described in Section 5.4. Quoted 95th percentile and RMS error quantities are over the entire data set (i.e., for both float and fixed epochs). PV denotes the availability of an aperture-test-validated fixed solution for conditioning in the primary filter. Pf denotes the false-fix rate as determined by an excursion of the primary filter beyond 30 cm from the ground truth when conditioned on fixed ambiguities.