A high-fidelity testbed for evaluating ambulatory augmented reality human-machine interfaces

As augmented reality (AR) systems are increasingly employed in high stakes contexts, evaluating how interface design impacts user performance and cognitive workload is gaining in importance. ARIADNE (Augmented Reality Interface Assessment for Dismounted Navigation Environments) is a high-fidelity virtual reality (VR) testbed developed to assess human-machine interface (HMI) effectiveness in complex navigation and monitoring tasks. ARIADNE integrates dismounted ambulatory navigation in immersive VR, gesture-based interfaces, and a simulated AR heads-up display (HUD), while capturing quantitative behavioral and physiological performance metrics. As a proof-of-concept, we used ARIADNE to compare task performance and workload with two different HUD designs: a Fixed HUD, which instantaneously followed head orientation, and a Lagged HUD, which included a 150 ms delay. Participants (N = 23) completed a scenario involving a primary navigational search task and two secondary tasks: a Detection Response Task (DRT) and a threat monitoring task. DRT accuracy showed significant effects of HUD condition, session order, and their interaction while no significant differences emerged between HUD types in total navigation time, distance traveled, or threat detection accuracy. Our findings indicate that even brief delays in HUD responsiveness can measurably impact user performance, underscoring the importance of addressing latency in AR interface design. Importantly, this work demonstrates ARIADNE’s utility as a research platform capable of capturing fine-grained behavioral performance metrics, enabling systematic evaluation of AR system features in realistic, task-relevant contexts.