Enhancing Trimble Applanix POSPac INFusion+ PP-RTX Performance in Urban Canyons with a Novel GNSS Outlier Detection Method

Jau-Hsiung Wang

Peer Reviewed

Abstract:	Trimble Applanix’s Positioning and Orientation System Post-Processing (POSPac) is the industry-leading software suite that uses GNSS and inertial technologies to provide direct georeferencing for mobile mapping sensors across air, land, and marine platforms. Released in 2025, Trimble ProPoint single base Real-Time Kinematics (RTK) has been integrated into POSPac to enhance georeferencing robustness in GNSS-challenging environments by leveraging modern GNSS signals and advanced filtering. When a dedicated base station is impractical, Trimble CenterPoint® Real-Time eXtended (RTX) offers a valuable alternative. However, RTX performance can degrade significantly in urban canyons where multipath, signal blockage, and interference introduce multiple measurement outliers and lead to unreliable solutions. These degraded RTX outputs often underestimate their uncertainties, destabilizing GNSS-inertial integration and causing inertial navigation jumps. Existing GNSS outlier detection approaches such as Receiver Autonomous Integrity Monitoring (RAIM), fading carrier-to-noise based checks, and methods relying on external sensors or 3D maps struggle when outliers occur simultaneously or when additional data sources are unavailable. This paper presents a new GNSS outlier detection method designed to improve IN-Fusion+ Post-Processing RTX (PP-RTX) performance in harsh urban environments without relying on external data. The method monitors pseudorange or code biases and constructs a per-satellite bias model from benign GNSS conditions. When operating in urban canyons, multipath and interference distort the computed code biases. The deviation between measured biases and the established model is used to estimate multipath and positioning errors. These estimates enable effective detection of large-error RTX solutions and improved RTX error modeling within the GNSS-inertial fusion filter. The method has resulted in two patent applications. The approach was evaluated using over 65-hours of downtown Toronto datasets collected with five Trimble Applanix mobilemapping systems. Reference trajectories were generated using tightly coupled Trimble ProPoint single base RTK with a navigation grade IMU. Results show that the proposed method identifies major RTX outliers and reduces PP-RTX position error by 87.84% at 1? and 94.39% at 2?. Incorporating the method into IN-Fusion+ PP-RTX improves post-processed position accuracy by about 44.37% at 1? and 57.92% at 2?.
Published in:	Proceedings of the ION 2026 Pacific PNT Meeting April 13 - 16, 2026 Hilton Waikiki Beach Honolulu, Hawaii
Pages:	216 - 229
Cite this article:	Wang, Jau-Hsiung, "Enhancing Trimble Applanix POSPac INFusion+ PP-RTX Performance in Urban Canyons with a Novel GNSS Outlier Detection Method," Proceedings of the ION 2026 Pacific PNT Meeting, Honolulu, Hawaii, April 2026, pp. 216-229. https://doi.org/10.33012/2026.20627
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In