Supporting the Contact Tracing Process with WiFi Location Data: Opportunities and Challenges

Title of the Paper

Supporting Contact Tracing Processes Using WiFi Location Data: Opportunities and Challenges

Paper Information

• Subject Area: Human-Computer Interaction Technology and Public Health Systems
• Keywords: Contact tracing, WiFi data, visualization, human-computer interaction, data privacy, collaborative decision-making, pandemic response

Research Background and Problem

• Problems or Challenges:
  • Manual contact tracing relies on the memory of community members, which may be unreliable (e.g., misremembering or forgetting).
  • The rapid spread of COVID-19 and the high proportion of asymptomatic carriers pose significant challenges to existing contact tracing methods.
  • While digital support can accelerate detection and notification, it still faces issues of imperfect data (such as false positives or false negatives).
• Importance:
  • Contact tracing is critical for curbing the spread of infectious diseases and is a key component of non-pharmaceutical interventions (NPI).
  • Rapid and large-scale monitoring of virus transmission is crucial for public health system responses.
• Research Motivation and Related Work:
  • Exploring how technology can improve memory-related issues in manual contact tracing processes.
  • While digital contact tracing methods have been extensively studied, there is insufficient research on integrating digital tools into manual tracing processes.
  • This study aims to explore the potential of optimizing contact tracing using university campus WiFi data.

Solution

• Method or Solution:
  • Developed a contact tracing visualization tool based on campus WiFi network data.
  • The tool uses de-identified WiFi data to assist contact tracers in identifying and correcting memory errors during simulated contact tracing calls.
  • The tool's design was optimized based on user research and feedback, emphasizing privacy protection and intuitive data presentation.
• Innovations:
  • Introduced a digital technology leveraging university WiFi data to support the contact tracing process.
  • Designed a more user-friendly and de-identified control interface to help contact tracers effectively handle imprecise data.
  • The tool integrates filtering and sorting functions for time, location, and co-located collaborators' data, supporting more efficient collaborative decision-making.
• Implementation Steps:
  1. Obtain de-identified data from WiFi authentication logs, including access point identifiers, device identifiers, and timestamps.
  2. Develop a visualization interface: present shared time periods of community members and co-located individuals on a timeline.
  3. Optimize the design: conduct user testing and feedback collection on the initial design, adjust privacy protection measures, and add features such as date input fields and data filtering.
  4. Conduct simulated tracing experiments: perform simulated phone tracing tasks using the WiFi data visualization tool.
  5. Analyze the tool's effectiveness through phone tracing and interviews, and explore participants' perspectives.

Research Findings

• Specific Findings:
  • The visualization tool effectively helped contact tracers fill memory gaps of community members, significantly improving the efficiency of the tracing process.
  • Confirmed the usability of WiFi data and its benefits in verifying real-life events and location data.
  • Contact tracers previewed data using the tool before calls, enhancing their preparation and confidence for the task.
• Advantages Over Existing Solutions:
  • Compared to manual contact tracing, which relies solely on community members' memory, the visualization tool is more intuitive and user-friendly, actively detecting and correcting data discrepancies.
  • Does not require community members to install applications, lowering the barrier to technology adoption.
• Experimental or Evaluation Results:
  • In 14 simulated tracing calls, an average of 11 close contacts were identified per call.
  • Approximately 72% of community member data events had inconsistencies, but 96% of these were resolved through the tool and user collaboration.
  • The tool highlighted common memory issues in contact tracing scenarios and the importance of visualizing data.
• Limitations and Future Directions:
  • The tool relies on campus WiFi infrastructure, which may face data capture issues and omissions in non-campus environments or for users with low connectivity.
  • Privacy concerns among community members persist, and establishing greater trust mechanisms remains a challenge for the future.
  • Future research is recommended to explore design methods for addressing false positives in data and to test whether community members can directly access the tool to promote transparency and data awareness.

Conclusion and Design Recommendations

• Core Design Principle: Use the visualization tool as an auxiliary aid in the contact tracing collaboration process, rather than a complete replacement for decision-making.
• Key Design Principles: Support data prioritization and filtering, flag potential inconsistencies, and enhance user data adjustment capabilities.
• Privacy and Community Engagement: Implement clear communication strategies about data collection and usage to alleviate community concerns about data breaches.