Exploring Spatial UI Transition Mechanisms with Head-Worn Augmented Reality

Document Title

Exploring Transition Mechanisms for Spatial User Interfaces in Head-Mounted Augmented Reality

Document Information

• Subject Area: Human-Computer Interaction, Augmented Reality, Interface Design
• Keywords: Augmented Reality, Automation, Controllability, Adaptive Interface, User Experience, Error Prediction, Spatial User Interface

Research Background and Issues

• Identified Problems or Challenges:

  • In current augmented reality (AR) operating systems, interface content is typically fixed in one position, requiring users to manually adjust or move the UI, which increases workload.
  • Users' task information needs change across different spaces, but existing mechanisms fail to meet these dynamic demands.
  • Automatically predicting and arranging interface content may lead to prediction errors, and mitigating their negative impact on user experience remains a challenge.

• Importance:

  • As AR becomes increasingly integrated into daily life, providing users with dynamic and adaptive interfaces can enhance task efficiency and reduce operational burden.
  • Human-computer interaction technology is evolving, necessitating the design of efficient, user-centric, low-friction interaction mechanisms.

• Research Motivation and Related Work:

  • Previous studies have explored the potential of information display in AR, but most focus on static, fixed UI layouts rather than dynamic applications in user environments.
  • When addressing dynamic interaction, balancing system automation, user controllability, and error handling is a critical issue.

Solution

• Proposed Solution or Method:

  • This study designs three UI transition mechanisms with varying degrees of automation: low operational burden (Wristpack), semi-automatic (Semi-Auto), and fully automatic (Fully-Auto).
  • It explores how to integrate automation and controllability in spatial interactions while simulating the impact of prediction errors on user experience.

• Innovative Aspects:

  • Introduced three interface solutions combining automation and user control, and conducted comparative experiments to reveal their performance differences.
  • Systematically examined the cost of prediction errors on user experience for the first time, proposing design recommendations for low-cost error recovery.

• Implementation Steps and Key Techniques:

  • Design Workshop: Captured user pain points in AR interface usage through practical tasks and proposed potential solutions.
  • Experimental Platform: Used VR to simulate real environments, addressing AR device limitations such as restricted field of view and unstable multi-room tracking.
  • Experiment: Evaluated four UI transition mechanisms—Wristpack, Semi-Auto, Fully-Auto, and a traditional baseline method—in a virtual home scenario.
  • Data Collection and Analysis: Assessed efficiency, usability, workload, user preference, and error recovery efficiency.

Research Results

• Specific Findings:

  • The Semi-Auto mechanism performed best in terms of efficiency (task completion time and movement distance), usability, workload, and user preference.
  • Users experienced significantly lower control with the Fully-Auto mechanism compared to manual operation, while the Semi-Auto mechanism effectively maintained high user control.
  • In cases of prediction errors, the Semi-Auto mechanism had significantly shorter error recovery times than the Fully-Auto mechanism.

• Comparison with Existing Solutions:

  • Compared to traditional manual dragging and resetting UI baseline methods, the three proposed mechanisms significantly reduced operational burden.
  • While Fully-Auto had the highest level of automation, its error recovery cost was high; in contrast, Semi-Auto struck a balance between user controllability and automation, successfully enhancing user experience.

• Experimental or Evaluation Results:

  • Time and Efficiency: The baseline method required the longest time to complete tasks, while Semi-Auto was the most efficient.
  • Accuracy: Semi-Auto simulated an effective prediction system with low error recovery costs, resulting in higher accuracy.
  • Subjective Feedback: Due to its high controllability, the Semi-Auto mechanism was the most favored by users.

• Limitations and Future Directions:

  • Limitations:
    • The experiment was based on VR simulation and not applied in real AR device scenarios.
    • Navigation methods in virtual environments (e.g., teleportation) may affect simulation accuracy and user behavior.
    • Prediction accuracy was fixed at 75%, and further exploration of accuracy impacts in other systems is needed.
  • Future Directions:
    • Validate the mechanisms in real AR environments.
    • Explore lightweight error recovery methods to balance efficient automation and controllability.
    • Investigate UI requirements in diverse user task environments, such as non-urgent tasks and non-tool-related information displays.

Research Significance

By combining automation and user control, this study addresses the issue of AR interface transitions in complex dynamic task scenarios, offering new insights into designing low-operational-burden augmented reality user interfaces while emphasizing the impact of prediction errors on user experience.