Barometer-Driven Height Estimation for Indoor Positioning Including Elevator and Escalator Detection Mechanisms

Philipp Hager, Susanna Kaiser, and Christian Gentner

Peer Reviewed

Abstract:	The field of infrastructure-free indoor location estimation for pedestrians is a highly active area of research. A promising approach involves using foot-mounted inertial sensors combined with step-wise re-calibration, achieving meter-level accuracy in three dimensions (3D). However, certain applications require more precise vertical accuracy, and a height error of a few meters can be problematic. In this paper, we show how the information of additional barometer sensors leads to a more accurate three-dimensional (3D)-position estimation and an overall more reliable positioning performance. Additionally, we classify the vertical motion, such as stair climb, elevator, escalator or mixed usage in parallel. To evaluate the performance of the proposed height estimation and classification algorithms, we conducted multiple indoor tests in various multi-level buildings. The height error is assessed using the known floor-to-floor distance while the classification accuracy is assessed by comparing the results of the classifier to manual labels. We compare the height error with and without the proposed height estimation algorithm, demonstrating that the proposed algorithm increases vertical positioning accuracy up to 75 % and is able to accurately classify vertical motion with 98 % overall accuracy. Index Terms—barometer, inertial measurement unit, state machine, classification
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:	1280 - 1288
Cite this article:	Hager, Philipp, Kaiser, Susanna, Gentner, Christian, "Barometer-Driven Height Estimation for Indoor Positioning Including Elevator and Escalator Detection Mechanisms," 2025 IEEE/ION Position, Location and Navigation Symposium (PLANS), Salt Lake City, UT, April 2025, pp. 1280-1288.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In