Can we use OpenStreetMap POIs for the Evaluation of Urban Accessibility

GIScience 2016 Short Paper Proceedings Can we use OpenStreetMap POIs for the Evaluation of Urban Accessibility? S. Steiniger 1 , M.E. Poorazizi 2 , D.R. Scott 1 , C. Fuentes 1 , R. Crespo 3 CEDEUS & Pontificia Universidad Catolica de Chile A, Avda. Vicuna Mackenna 4860, Macul - Santiago - Chile Email: ssteiniger@uc.cl University of Calgary, 2500 University Drive NW, Calgary, AB T2N 1N4, Canada Email: mepooraz@ucalgary.ca Universidad Bernardo O’Higgins, Avenida Viel 1497 Ruta 5 Sur - Santiago, Chile Email: ricardo.crespo@ubo.cl Abstract High urban accessibility, measured in terms of easy and fast to reach public and private services, such as schools, parks and shops, is considered an important indicator for quality of life. We developed a web-based platform to measure urban accessibility for transportation planning. Several cities that we aim to study are in Chile where data on urban services is scarce. We evaluated if OpenStreetMap (OSM) can be used as a source of Points of Interest (POI) to evaluate urban accessibility. Based on a field survey in Santiago de Chile, we found that completeness of the OSM POI database is geographical very diverse (7% - 73%) and therefore accessibility scores change significantly. However, scores in some areas did not change much when POIs were added to the OSM database. Given the lack of alternative data sources we recommend using OSM, but suggest evaluating POI completeness in areas with average and low accessibility scores. 1. Urban Accessibility Urban accessibility measures how easy or difficult it is to visit places of one’s day-to-day activities by measuring travel time. As daily activities are considered things such as going to work or school, visiting the library or a park, doing (grocery) shopping, going out to a restaurant, or visiting a movie theatre. Traditionally, transportation planning has focused on measuring accessibility as accessibility to work places, as work-related trips generate high travel demand. Following newer but also classic accessibility measurement theory and practice (e.g., Talen and Anselin 1998; Geurs and van Wee 2004), we developed a web-based platform that measures urban accessibility by counting activity opportunities within a certain travel time – or “travel-shed”. The platform should later be used to evaluate how accessibility changes when investments into urban infrastructure or changes to public transit schedules are made. To evaluate urban accessibility based on activity opportunities and with respect to different modes of travel (e.g. walking, biking, public transit) different data sources are needed: (a) activity opportunities in form of Point of Interest (POI) data, (b) road and pedestrian path network information, and (c) bus and subway/LRT schedules if accessibility with public transit is of interest. In this paper we aim to evaluate how completeness of the (OpenStreetMap) POI database affects calculated walkability scores. 2. Why OpenStreetMap as a Data Source Our first platform prototype was implemented to measure accessibility in Calgary, Canada (Steiniger et al. 2013). For this prototype we used road network data from OpenStreetMap and Points of Interest (POI) were received from MapQuest’s proprietary online database.