“I can run at night!": Using Augmented Reality to Support Nighttime Guided Running for Low-vision Runners

Research Background and Issues

• Issues and Challenges: Individuals with low vision (LV) face visual limitations when running at night, which not only hinders their ability to utilize residual vision for environmental recognition and following guidance but also increases the risks associated with nighttime exercise. Insufficient lighting prevents them from clearly perceiving their surroundings, further restricting their ability to engage in Caller-style guided running at night.
• Significance: Running is vital for physical, mental, and social well-being. However, LV individuals are unable to safely and effectively participate due to visual impairments, especially at night, depriving them of crucial exercise opportunities. This impacts physical health and may exacerbate psychological stress.
• Research Motivation: Existing augmented reality (AR) technologies have been applied to support LV individuals in indoor activities (e.g., navigating stairs, indoor mobility, and cooking), but their applicability in unstructured, low-light outdoor environments remains unclear. Nighttime running, in particular, involves complex environmental factors such as terrain changes and dynamic obstacles (e.g., vehicles or other runners).

Solution

• Method and Solution: A novel AR-based assistive tool called "RunSight" is proposed to support Caller-style nighttime running. Utilizing a see-through head-mounted display (HMD) and image enhancement technologies, RunSight provides enhanced visual perception while visualizing the guide’s position through AR.
• Innovations:
  • Introducing AR technology into the nighttime outdoor running scenarios of LV individuals, marking a new application domain.
  • Combining multiple AR visualization methods (e.g., edge enhancement EDGE, guide position indication GUIDE AR) through iterative design to meet diverse LV user needs.
  • Employing ultra-wideband (UWB) technology for real-time guide positioning, ensuring precise direction and distance cues while integrating voice communication between the user and guide.
• Implementation Steps:
  1. Initial Design and Prototype Development: Two rounds of iterative design explored the effectiveness of edge-enhanced visual cues (EDGE) and auditory cue technologies.
  2. Integration of AR Technology: Expanded into the final RunSight prototype, incorporating GUIDE AR multi-mode options (e.g., POINT, ARROW, and RAILROAD) to assist with target positioning.
  3. User Research and Optimization: Conducted running tests and gathered feedback from 8 LV runners to continuously refine system usability and performance.

Research Outcomes

• Specific Results:
  • Nighttime Running Support: All 8 participants successfully completed at least 1 kilometer of nighttime running with RunSight assistance (average running distance: 3.44 kilometers), whereas none could safely run without RunSight.
  • Guidance Effectiveness: LV individuals effectively followed the guide, maintaining a stable guide-runner distance (average distance: 2.2 meters, standard deviation: 0.45).
• Advantages Compared to Existing Solutions:
  • Existing solutions are mostly limited to daytime activities or simple track scenarios, while RunSight is the first to enable nighttime outdoor running in low-light, complex environments.
  • The integration of EDGE enhancement and GUIDE AR guidance provides more comprehensive support compared to single auditory or tactile cues.
• Experiments and Evaluation:
  • Quantitative Results: Participants’ psychological sense of safety significantly improved (average score exceeding 5), and most expressed positive attitudes toward its social acceptability (scores ranging from 6 to 7).
  • Subjective Feedback: Participants generally praised the enhanced visual field and guidance effects but identified the weight of the head-mounted device (HMD) as the primary physical limitation.
  • Behavioral Analysis: Voice communication and AR assistance complemented each other, providing LV individuals with comprehensive perceptual support.
• Limitations and Future Directions:
  • Device Limitations: The current HMD weighs over 1 kilogram, limiting comfort during use. Future research should explore lighter devices with wider fields of view.
  • Guide Variability: The study’s guides were experienced running instructors, necessitating further validation of RunSight’s applicability with ordinary family members or unfamiliar guides.
  • User Diversity: The study focused on male participants; future research should include more diverse gender and visual ability groups to further verify its generalizability.
  • Advanced Application Scenarios: Future tests should explore RunSight’s potential in more challenging environments, such as uneven terrain, long-distance running, or marathons.

This study provides a feasible solution for nighttime running among LV individuals and offers an innovative direction for utilizing AR technology to enhance spatial perception and navigation capabilities in low-light environments.