Development of Pedestrian Level of Service (PLOS) model and satisfaction perception rating models for pedestrian infrastructure for mixed land-use urban areas

Existing pedestrian facilities are evaluated using a range of Pedestrian Level of Service (PLOS) assessment models. These models, available for well-defined facilities, consider factors like pedestrian volume and speed, physical infrastructure conditions and perceived satisfaction levels for comfort, safety and security of the facility. The satisfaction level, usually assessed using sample satisfaction survey, is subjective. No comprehensive PLOS assessment model conceptualized in this work is available for old cities having mixed land-use where well-defined pedestrian facilities might not have usually existed. This study eventually identifies relevant PLOS assessment factors for mixed land-use urban areas from literary narratives/previous research works and develops a comprehensive PLOS assessment model for them considering all these factors. The factors were initially grouped into factor groups or parameters using principal component analysis done with importance survey responses from 550 pedestrians from Patna and Gaya, two old cities of India. Six important parameters have been identified namely safety issues under pedestrian traffic interaction; condition of pedestrian infrastructure; pedestrian convenience and sense of security; night time walking; encroachment and walking comfort. The model considers pedestrians’ satisfaction for the parameters which is a function of actual conditions for the underlying factors and varies from person to person. To eliminate variability, the study develops pedestrian satisfaction rating models for each parameter through satisfaction survey, correlating existing pedestrian facility condition to perceived satisfaction level, using ordered probit model. Random conditions for each parameter was designed through D-Optimal experimental design considering four levels (best to worst) of factor conditions and survey was done with 780 participants, each participating in 16 experiments.