Measuring the Accuracy of Inside-Out Tracking in XR Devices Using a High-Precision Robotic Arm

Daniel Eger Passos, Bernhard Jung

Abstract: We present a method for measuring the accuracy of inside-out tracking capabilities of XR devices. The XR device is attached to an industrial robotic arm that can repeat motions with high precision. A calibration procedure based on point cloud matching is used to determine the relative transformation between the robot arm and the XR device. In tests conducted so far, we experimented with different XR devices, and lighting conditions. For example, under good environmental conditions, tracking accuracies of $${<}$$ 1 cm were achieved by the Oculus Quest and $${<}$$ 2 cm by the Samsung Galaxy S9. However, under less benevolent environmental conditions, mean error and variance increased significantly. We conclude that the proposed method provides high repeatability of conducted experiments. It also offers diverse opportunities for future investigations regarding the sensitivity of achievable tracking accuracies of XR devices in different environment conditions such as lighting and feature richness.