Assessing the Accuracy of an Outdoor Augmented Reality Solution for Mapping Underground Utilities

Abstract

Augmented reality (AR) has been identified as a technique that can enhance information extraction from the virtual world to the real world and improve the access and utilization of information. The main challenge in creating a reliable and adaptive outdoor AR system is the accurate registration of virtual objects in the real world. There is currently limited research using outdoor AR for collecting underground utility inventory and evaluating the accuracy. Because of the limited accuracy of smartphones, this study used external GPS devices to reduce positional error. The main objective of this research was to evaluate the horizontal accuracy of the captured data used by the AR phone application XR-GIS, developed by the authors. This study used root-mean-square error (RMSE), average Euclidean error (AEE), and central error (CE) in the evaluation of two smartphones and two popular GPS devices on 16 survey location points. Leveraging a Trimble S5 robotic total station, a geodetic base point and the location points used in the study were authoritatively measured. The results show that the smartphone accuracies were significantly different compared with smartphones linked to external GPS devices. Generally, the results suggest that current smartphones cannot be adopted in the underground construction industry without higher-accuracy integrated GPS devices.