Robust Railway Ground Truth and Confidence Bounds Estimation Based on RTK/INS Forward-Backward Smoothing

Anja Grosch and Omar García Crespillo

Peer Reviewed

Abstract:	Accurate and robust trajectory determination, incorporating a reliable quality metric, is imperative for railway ground truth determination and digital map generation. A significant number of international projects have previously been undertaken in the field of robust and suitable railway localization. However, a uniform reference/ground truth computation has not been employed by any of these projects, and the absence of any explicit confidence information regarding these solutions hinders the capacity to perform a meaningful performance comparison of the various proposed methods. It is therefore proposed that a methodology be established for the determination of trajectories and the associated confidence intervals, with the objective of achieving both high accuracy and reliability. The methodology employed is predicated on the utilization of three sensors: GNSS, IMU and RTK base station, with the objective of acquiring the required trajectory quality. The GNSS carrier-phase positioning is achieved through a tightly coupled RTK/INS fusion framework, complemented by two-filter forward and backward smoothing techniques, to ensure the attainment of a high-accuracy solution. In order to obtain a robust and reliable trajectory determination, a multi-layer fault detection and exclusion architecture is applied with the objective of reducing the likelihood of sensor faults, particular GNSS faults, entering and propagating through the processing steps. In addition, a multi-hypothesis solution separation approach is utilized for the purpose of validating the results and providing the desired confidence intervals. The present study is based on a railway measurement campaign that was conducted in Malaga, Spain. The results of this campaign demonstrate ´ that high accuracy in trajectory determination can be achieved. Furthermore, it is shown that the horizontal confidence interval reliably overbounds the position error. Index Terms—RTK/INS, Forward-Backward Smoothing, Fault Detection and Exclusion, Multi-Hypotheses Solution Separation, Confidence Interval, Ground Truth
Published in:	2025 IEEE/ION Position, Location and Navigation Symposium (PLANS) April 28 - 1, 2025 Salt Lake Marriott Downtown at City Creek Salt Lake City, UT
Pages:	1336 - 1347
Cite this article:	Grosch, Anja, Crespillo, Omar García, "Robust Railway Ground Truth and Confidence Bounds Estimation Based on RTK/INS Forward-Backward Smoothing," 2025 IEEE/ION Position, Location and Navigation Symposium (PLANS), Salt Lake City, UT, April 2025, pp. 1336-1347.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In