Geo-LSTM: A Geometry and Temporal Feature Fusion Algorithm for Multi-Sensor 3D Localization
Kai Li, Le Bao, Wansoo Kim
AI summary
Problem
High-precision 3D localization for human-robot collaboration in dynamic indoor spaces is hindered by UWB signal multipath, occlusions, and the lack of reliable vertical accuracy in existing wireless ranging systems.
Approach
The method combines a lightweight Simplified Geometric Localization algorithm with a dual-branch LSTM network to extract temporal features from both UWB ranging data and multi-sensor-derived a priori coordinates.
Key results
- Achieves 0.103 m average 3D localization RMSE in cluttered indoor environments
- Improves accuracy by 38.60% over weighted least squares and 31.20% over range-based LSTM
- Reduces computational complexity to O(1) per time step via the SGL algorithm
- Maintains robust performance under both line-of-sight and non-line-of-sight conditions
Why it matters
Provides a reliable, low-cost localization solution critical for safe and adaptive human-robot collaboration in real-world indoor applications.
Abstract
Accurate three-dimensional (3D) localization is crit- ical for robust human-robot collaboration (HRC) in dynamic indoor environments. However, realizing high-precision local- ization in complex scenarios still faces challenges such as multipath effects, field-of-view occlusion, etc. To address these limitations, we propose Geo-LSTM, a geometry-constrained long short-term memory (LSTM) framework that integrates ultra- wideband (UWB) sensors, inertial measurement unit (IMU), and barometric pressure (BMP) sensors. First, a Simplified Geometric Localization (SGL) algorithm is proposed, which uses dual-BMP sensors and IMU sensor to obtain precise height information and utilizes the geometric relationships between the UWB tag and anchors to compute an initial location estimate, serving as a priori input for the Geo-LSTM network. This Geo-LSTM algorithm then incorporates multi-source geometric information to extract time-series features from the UWB ranging data and the tag’s a priori location, further enhancing 3D localization accuracy. The experimental results from the cluttered indoor environments, including real-world HRC tasks with occlusions, show that the Geo-LSTM algorithm achieves an average 3D local- ization root mean square error (RMSE) of 0.103 m, representing improvements of 38.60% and 31.20% over the weighted least squares (WLS) method and the range-based LSTM algorithm, respectively. These results demonstrate Geo-LSTM’s potential for reliable multi-sensor 3D localization in HRC applications.