Transcript – Crowd-Biosensing of Location-Based Physical and Emotional Distress for Walkable Built Environment

The conditional built environment in urban spaces can negatively impact the physical well-being and satisfaction of urban citizens. So, assessing and managing a diverse element of the built environment is important for promoting our quality of life. One promising method to do that is using physiological signals. Listen, development in wearable technology allow us to collect various physiological signals such as gay patterns, electric motor activity and heart rate such as signals can provide insight into how people feel and how people react to surrounding environments beyond the self reports. Therefore, our research team conducted several experiments to examine the usefulness of physiological signals. During the experiment, participants were requested to walk, predefine the path, and they asked to wear the wearable sensors like sensors, smartphone and inertial measurement unit. The collected physiological signals could capture the environmental stimuli over a short period of time, but signal noises and artifacts make it hard to identify the stimulus in entire data streams. So we proposed a of data processing technique for surveillance detection analysis to mitigate the exposure of uncontrollable confounding factors. This technique shows promising results in identifying the effects of environmental stimuli such as broken house and sidewalk defects. Additionally, we examine the interplay of multimodal data not only from the physiological signals but also from the image based data. Such multimodal data could deliver supplementary information to machine learning model, enhancing each ability to correctly classify the environmental stimuli. Currently, we are preparing to test this method in daily life settings. We expect that this approach and outcome will provide opportunities for advancing urban built environment assessments. Thank you.