Location-based collective distress using large-scale biosignals in real life for walkable built environments

Area of Science:

• Environmental Health
• Urban Planning
• Biomedical Engineering

Background:

• Wearable sensors offer objective measurement of pedestrian distress from negative environmental stimuli.
• Traditional surveys have limitations in reliability and subjectivity for assessing environmental distress.
• Prior research primarily used controlled settings, limiting real-world applicability.

Purpose of the Study:

• To investigate the usability of real-world biosignals for capturing pedestrian environmental distress.
• To correlate physiological responses with self-reported negative stimuli in ambulatory settings.
• To develop a data-driven approach for assessing urban environmental quality.

Main Methods:

• Collected geocoded biosignals (electrodermal activity, gait, heart rate) and self-reported stimuli data in real-life settings.
• Employed spatial analysis, statistical modeling, and machine learning techniques.
• Utilized machine learning to predict location-based collective distress.

Main Results:

• Machine learning model achieved 80% accuracy in predicting pedestrian distress.
• Demonstrated statistically significant associations between biosignals and environmental stimuli.
• Validated the use of ambulatory biosignals for environmental distress detection.

Conclusions:

• Real-world biosignals are effective for assessing pedestrian environmental distress.
• This approach enhances objective evaluation of built environments and urban dynamics.
• Integrates physiological data into walkability assessments and urban planning applications.