From Invisible to Actionable: Augmented Reality Interactions with Indoor CO2

Paper Title

From Invisible to Actionable: Augmented Reality Interactions with Indoor CO2

Publication Info

• Topic area: Augmented reality for environmental awareness and indoor air quality management.
• Keywords: Augmented reality, indoor air quality, CO2 visualization, wearable sensors, environmental awareness, user engagement, pollution mitigation, human-computer interaction, gamification, real-time monitoring.

Background and Problem

• Problem / challenge: Indoor CO2 pollution is invisible, odorless, and often overlooked despite its significant health risks. Existing static sensors are limited in providing actionable, spatially dense, and personalized insights for occupants.
• Significance: With people spending up to 90% of their time indoors, addressing indoor air quality is critical for health and cognitive performance. Effective visualization and actionable strategies are needed to mitigate CO2 hotspots.
• Motivation and related work: Prior research has focused on static pollution sensors and outdoor air quality visualization but lacks solutions for personalized, interactive, and actionable indoor pollution management. Augmented reality (AR) has shown potential for environmental awareness but remains underexplored for indoor pollutants like CO2.

Solution

• Proposed approach: The paper introduces CoWear, a wrist-worn CO2 sensor integrated with a smartphone-based AR application that visualizes CO2 concentrations as interactive, color-coded bubbles. The system enables users to monitor and mitigate CO2 hotspots in real time.
• Novelty:
  1. Development of a wearable CO2 sensor (CoWear) for personalized, real-time air quality monitoring.
  2. Integration of AR visualizations to represent CO2 concentrations as spatially anchored, interactive bubbles.
  3. Introduction of a gamified AR interface to engage users in pollution awareness and mitigation strategies.
• Procedure and key techniques:
  1. Design and implementation of CoWear with CO2, temperature, and humidity sensors, offering a 5-second sampling interval and wireless connectivity.
  2. Development of an AR app that visualizes CO2 levels as bubbles with varying size and color, representing pollution severity.
  3. User experiments in semi-controlled and in-the-wild settings to evaluate system usability, effectiveness, and user engagement.

Results

• Concrete findings:
  • Average CO2 reduction of 558 ppm using the AR app, compared to 201 ppm with a 2D heatmap.
  • Median usability score of 1.88 on the Post-Study System Usability Questionnaire (PSSUQ), indicating strong practicality.
  • Participants reduced CO2 by 479 ppm in session S1 and 283 ppm in session S2, with faster ventilation in subsequent sessions.
• Advantage over baselines:
  • AR visualization outperformed static 2D heatmaps in enabling targeted ventilation strategies and reducing CO2 levels.
  • Higher user engagement and awareness, with 90.3% of participants willing to recommend the system to others.
• Experiments / evaluation:
  • Conducted with 35 participants across semi-controlled (office rooms) and in-the-wild (home, diner, lab) settings.
  • Evaluated user competence, immersion, and positive experiences using Game Experience Questionnaire (GEQ) and PSSUQ.
• Limitations and future work:
  • Responsiveness delay (~30 seconds) in updating CO2 values.
  • Potential clutter in AR visualization due to overlapping bubbles.
  • Plans to integrate additional pollutants (e.g., VOCs, PM2.5), multiplayer features, and context-aware recommendations.

Summary

This paper presents CoWear, a wrist-worn CO2 sensor paired with an AR application that visualizes indoor CO2 levels as interactive bubbles, enabling users to identify and mitigate pollution hotspots. The system demonstrated significant CO2 reductions (average 558 ppm) and improved user awareness and engagement compared to traditional 2D heatmaps. With a median usability score of 1.88, the approach shows promise for real-world applications in promoting healthier indoor environments. Future work aims to address responsiveness, expand pollutant coverage, and incorporate collaborative features.