Level of service criteria of urban walking environment in indian context using cluster analysis

To know how well roadways accommodate pedestrian travel or how they are pedestrian friendly it becomes necessary to assess the walking conditions. It would also help evaluating and prioritizing the needs of existing roadways for sidewalk construction. Estimation of Pedestrian Level of Service (PLOS) is the most common approach to assess the quality of operations of pedestrian facilities. The focus of this study is to identify and access the suitable methodology to evaluate PLOS for off-street pedestrian facilities in Indian context. Defining the level of service criteria for urban off-streets pedestrian facilities are basically classification problems. Cluster analysis is found to be the most suitable technique for solving these classification problems. Cluster analysis groups object based on the information found in the data describing their relationships. K-means, Hierarchical Agglomerative Clustering (HAC), Fuzzy c-means (FCM), Self Organizing MAP (SOM) in Artificial Neural Network (ANN), Affinity Propagation (AP) and Genetic Algorithm Fuzzy (GA Fuzzy) clustering are the six methods are those employed to define PLOS criteria in this study. Four parameters such as pedestrian space, flow rate, speed and volume to capacity (v/c) ratio are considered to classify PLOS categories of off-street pedestrian facilities. And from the analysis six LOS categories i.e. A, B, C, D, E and F which are having different ranges of the four parameters are defined. From the study it found that pedestrian faces a good level of service of “A”, “B” and “C” are more often than at poor levels of service of “D”, “E” and “F”. From all the six clustering methods K-means is found to be the most suitable one to classify PLOS in Indian context