Return to Session D6b Next Abstract

Session D6b: Smartphone-Based Localization

A Method to Estimate User Position and Orientation Based on Relative Position Measurement Between a Navigation Robot and a Smartphone
Satoki Ogiso, National Institute of Advanced Industrial Science and Technology (AIST); Daisuke Sato, Carnegie Mellon University; Ryosuke Ichikari, Takahiro Miura, Masakatsu Kourogi, Takashi Okuma, Takeshi Kurata, and Takeshi Tanabe, National Institute of Advanced Industrial Science and Technology (AIST)
Location: Grand Ballroom ABC
Date/Time: Thursday, May. 1, 4:05 p.m.

This paper proposes a method to estimate a smartphone’s location and orientation using the relative position between it and a navigation robot. The relative position is measured by ultra-wideband (UWB) ranging and angle of arrival (AoA). The smartphone and navigation robot estimate their positions in their coordinate systems. The proposed method estimates the transformation of these two coordinate systems using UWB measurements. The localization error was evaluated by an experiment in a museum with six participants. As a result, the localization error of the proposed method was 0.40 m in the median, which is better than 0.65 m, the error with a baseline method that aligns coordinates at the start of localization. This method enables higher accuracy of smartphone localization for the users of navigation robots in short-term leave.
Index Terms—Ultra-wideband, Visual-inertial odometry, Navigation robot, Relative position, Extended Kalman filter

Return to Session D6b Next Abstract