Development of pedestrian safety index models for safety of pedestrian flow at un-signalized junctions on urban roads under mixed traffic conditions using MLR

Abstract

In metropolitan areas of emerging countries, such as India, there has been a fast rise in the number of pedestrians and traffic clashes in recent decades. In order to improve the safety and security of pedestrian services under heterogeneous traffic conditions, it is important to improve pedestrian facilities along city roads. The research study was undertaken to develop Pedestrian Safety Index models using parameters such as crosswalk speed (m/sec), crosswalk markings, vehicle stream speed (km/hr), pedestrian volume, pedestrian safety scores /rankings and crosswalk length (m). Data have been collected and extracted through videographic technique at various un-signalized junctions in the Central Business Districts (CBD) zones and surroundings of Srinagar Metropolitan (SMA) Area, where increase in the number of pedestrian volumes was identified. The amount of pedestrian volumes and average pedestrian speed (m/sec) during peak hours were identified at some of these study locations. Pedestrians have been identified on the basis of different age categories. Pedestrian movements are classified according to the speed and direction of the crossing. The Pedestrian safety index models are developed using SPSS software and tested on dataset obtained at various locations in the Srinagar city. The developed Pedestrian Safety Score Index (PSSI) model results obtained for all selected locations in the city of Srinagar are three to four indicating the ordinary / average safe to danger situation of the PSI scores. The results of the PSSI and Ped ISI model demonstrate the estimation of the correct index value of pedestrian safety at various un-signalized junctions. This research will also help to enhance the pedestrian safety strategies, pedestrian volumes (ped/hr) and also improved pedestrian safety interventions on urban roads under heterogeneous traffic environments.