Modeling and Evaluating the Impact of Mobile Usage on Pedestrian Behavior at Signalized Intersections: A Machine Learning Perspective

Pedestrian safety is a growing global concern, particularly in urban areas, where rapid urbanization and increased mobile device usage have led to an increase in distracted walking. This study investigates the impact of technological distractions, specifically mobile usage (MU), on pedestrian behavior and safety at signalized urban intersections. Data were collected from 11 signalized intersections in New Delhi, India, using video recordings. Key inputs to the modeling process include pedestrian demographics (age, gender, group size) and behavioral variables (crossing speed, waiting time, compliance behaviors). The outputs of the models focus on predicting mobile usage behavior and its association with compliance behaviors such as crosswalk and signal adherence. The results show that 6.9% of the pedestrians used mobile phones while crossing the road. Advanced machine learning models, including Convolutional Neural Networks (CNN), Long Short-Term Memory networks (LSTM), and Recurrent Neural Networks (RNN), have been applied to analyze and predict MU behavior. Key findings reveal that younger pedestrians and females are more likely to exhibit distracted behavior, with pedestrians crossing alone being the most prone to mobile usage. MU was significantly associated with increased levels of crosswalk violation. Among the machine learning models, the CNN demonstrated the highest prediction accuracy (94.93%). The findings of this study have a practical application in urban planning, traffic management, and policy formulation. Recommendations include infrastructure improvements, public awareness campaigns, and technology-based interventions to mitigate pedestrian distractions and to enhance road safety. These findings contribute to the development of data-driven strategies to improve pedestrian safety in rapidly urbanizing regions.