Abstract
Identifying and eliminating slip, trip, and fall (STF) hazards in the built environment is essential for preventing fall accidents. Wearable motion sensors and collective sensing-based approaches allow for continuous monitoring of STF hazards without manual inspection. This method works by monitoring the loss of balance (LOB) and analyzing fall risk scores. However, normal walking includes postural shifts like bending and hopping, which can mimic fall-like motion patterns, leading to inaccurate hazard identification. To address this, a wearable biosensor can be utilized, which can be a part of the same device with the motion sensor (e.g., wristband), as physiological arousal due to hazard exposure might be higher than during regular walking with postural shifts. The authors have developed a combined motion and biosensor-based method to identify STF hazards. The motion sensor monitors body movement while the biosensor measures physiological arousal. Fall risk is assessed by combining abnormal body movement scores with anomaly arousal scores using an unsupervised approach. A heatmap and Monte Carlo method are then applied to statistically identify locations with high fall risk. This approach was tested with 60 subjects on a predefined route. Results indicate the feasibility of the proposed technique in identifying STF hazards, thereby helping to develop strategies for hazard elimination.
Presenters
SangHyun LeeProfessor, Civil and Environmental Engineering, University of Michigan, Michigan, United States
Details
Presentation Type
Paper Presentation in a Themed Session
Theme
KEYWORDS
Fall risk in the built environment, Pedestrian safety
