Pedestrian overpass utilization modeling based on mobility friction, safety and security, and connectivity using machine learning techniques

Abstract

The identification, understanding, and treatment of predictors that drive the use of pedestrian overpass or Foot Over Bridge (FOB) are very much essential for city planners and policymakers. There is paucity of studies that uses modern soft computing techniques to understand the factors driving the use of FOBs. The aim of the work presented here was to identify the important predictors that determine usability of FOBs. The study utilized both questionnaire survey (perception in terms of satisfaction/dissatisfaction) and field data collected across fourteen locations in six different Indian cities. The goal was to identify the essential features that drive the usability of FOBs under four different contexts, i.e., mobility friction, safety and security, and vertical connectivity and horizontal connectivity. Three soft computing algorithms such as generalized linear model (GLM), random forest (RF), and gradient boosting machine (GBM) were trained to predict the future usability of pedestrians. The modelling approach involved data collection from 14 FOB locations, preprocessing involving data input in spreadsheets, removing missing values and normalization for model training. The next stage involved splitting the data into training and testing set, followed by model training and hyper-parameter optimization using tenfold cross-validation. Finally, the developed models were evaluated for test dataset for generalization. The study results revealed that GBM algorithm showed highest classification accuracy on test dataset over the other two techniques at various scenarios. GBM helped in identifying the essential parameters that drive the usability of FOBs under the four different contexts. Sensitivity analysis supported the fact that gender and age had significant impact on the choice of pedestrians under different contexts. Further, the respondents’ feedbacks regarding existing problems were used to validate the findings. The safety and security, walk environment, frequency of daily use, comfort, location type, length of travel, stairway dimensions and reduced walkable width affected the choice of using the FOBs. Therefore, provision of CCTV cameras and security personnel, removal of obstruction, provision of proper lighting and all-weather shade, and regular maintenance of the facilities will significantly improve the pedestrians’ choice to use the FOBs. The identification of important variables not only provides better insight of factors that affects the choice of pedestrians using the elevated facilities but also provides a valuable source of information to researchers, planners and policymakers to construct a better-planned pedestrian friendly infrastructure.