Estimating the Effects of Encumbrance and Walking on Mixed Reality Interaction

Research Background and Problem

• Problem Identification:
  The authors investigated the impact of situational constraints on user performance in mixed reality (MR) interactions, specifically focusing on the effects of "carrying weight" and "walking" on typical MR tasks. These constraints are common in real-world environments but have often been overlooked in earlier MR research.

• Significance:
  With the commercialization of mixed reality devices (e.g., Meta Quest 3, Apple Vision Pro), users are beginning to use these devices in real-world scenarios, such as in gyms, shopping, or driving. These situational constraints could significantly reduce interaction performance and user experience with MR devices, making it crucial to understand their impact to improve MR design.

• Motivation and Related Work:
  Previous studies have shown that environmental factors (e.g., cold, noise) negatively affect interaction performance on mobile and desktop computing devices. However, research on the impact of situational constraints in MR devices is still very limited. This study aims to provide new insights into this field, complementing the existing understanding of MR interactions.

Solution

• Methods and Approach:
  Through experimental analysis, the authors quantified the effects of "carrying weight" and "walking" on three typical MR tasks, including target acquisition (direct selection and ray-casting) and text entry. Using Bayesian regression models, they evaluated changes in metrics such as task completion time, error rate, pointing offset, and throughput.

• Innovations:

  1. The authors systematically studied the specific effects of carrying weight and walking on MR interactions for the first time.
  2. They proposed and validated a new regression model to quantify these situational constraints.
  3. They provided task design recommendations to minimize the impact of situational constraints.

• Implementation Steps and Techniques:

  1. Experimental Setup
     • Simulated "carrying weight" by attaching wristbands with different weights (0kg, 0.5kg, 1.0kg).
     • Simulated "walking" by setting a paced walking path.
     • The three tasks included target acquisition (direct selection and ray-casting) and text entry.
  2. Metric Measurement
     • For target acquisition tasks, metrics included movement time, pointing offset, error rate, and throughput.
     • For text entry tasks, metrics included throughput, uncorrected error rate (UER), and corrected error rate (CER).
  3. Data Analysis
     • Bayesian regression analysis was used to evaluate the specific impact of each constraint on the metrics and compare differences across tasks.

Research Findings

• Specific Findings:

  1. Impact of Carrying Weight
     • Carrying 1.0kg increased target selection time by 28% and reduced target acquisition throughput by 22%.
     • Text entry throughput decreased by 17%, while the uncorrected error rate increased by 50%.
     • Pointing offset was not significantly affected.
  2. Impact of Walking
     • Walking increased target selection time (by 63% for ray-casting).
     • Text entry throughput decreased by 51%, and error rates in ray-casting increased (8.4%).
     • Pointing offset and errors both showed significant increases.
  3. Combined Effects
     • Combining 1.0kg weight with walking increased ray-casting target selection time by 112%.
     • Text entry throughput decreased by 58%.

• Advantages and Significance:

  • Demonstrated that situational constraints significantly impact the performance of primary interfaces in mixed reality.
  • Provided critical data to support the design of MR systems that adapt to real-world environments.

• Experimental Evaluation and Validation:

  • Quantitative experiments using a 3×2 within-subject design tested trends in task metric changes.
  • User interviews provided qualitative data to validate experimental observations.

• Limitations and Future Directions:

  1. Limitations
     • The experimental conditions were simulated (wristband weights and fixed walking speeds), which may underestimate the complexity of real-world scenarios.
     • The study only examined basic tasks and did not cover complex MR operations in real-world contexts.
  2. Future Directions
     • Explore the effects of carrying weight, complex walking patterns, and more practical interaction tasks in real-world scenarios.
     • Investigate the potential of improved input methods (e.g., eye tracking, voice input) under situational constraints.
     • Enhance experimental design diversity, including more complex constraints (e.g., dynamic obstacles or carrying handheld objects).

This study enhances the understanding of MR interaction design in real-world scenarios and emphasizes the importance of considering situational constraints to improve user experience. These findings lay a solid foundation for designing MR systems capable of dynamically adapting to situational constraints in the future.