BEGIN:VCALENDAR VERSION:2.0 PRODID:-//pretalx//talks.osgeo.org//foss4g-2023//speaker//9C8HQM BEGIN:VTIMEZONE TZID:CET BEGIN:STANDARD DTSTART:20001029T040000 RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10 TZNAME:CET TZOFFSETFROM:+0200 TZOFFSETTO:+0100 END:STANDARD BEGIN:DAYLIGHT DTSTART:20000326T030000 RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3 TZNAME:CEST TZOFFSETFROM:+0100 TZOFFSETTO:+0200 END:DAYLIGHT END:VTIMEZONE BEGIN:VEVENT UID:pretalx-foss4g-2023-P7D3SC@talks.osgeo.org DTSTART;TZID=CET:20230628T152000 DTEND;TZID=CET:20230628T152500 DESCRIPTION:Travel time estimation is used for daily travel planning and in many research fields such as geography\, urban planning\, transportation engineering\, business management\, operational research\, economics\, hea lthcare\, and more (Hu et al.\, 2020). In public health and medical servic e accessibility studies it is often critical to know the travel time betwe en patient locations and health services\, clinics\, or hospitals (Weiss e t al.\, 2020). In support of a study aiming to characterize the quantity a nd quality of pediatric hospital capacity in the U.S.\, we needed to calcu late the driving time between U.S. ZIP code population centroids (n=35\,35 2) and pediatric hospitals\, (n=928) a total of over 32 million calculatio ns. There currently exist numerous methods available for calculating trave l time including (1) Web service APIs provided by big tech companies such as Google\, Microsoft\, and Esri\, (2) Geographic Information System (GIS) desktop software such as ArcGIS\, QGIS\, PostGIS\, etc\, and (3) Open sou rce packages based on program languages such as OpenStreetMap NetworkX (OS Mnx) (Boeing\, 2017) and Open Source Routing Machine (OSRM) (Huber & Rust\ , 2016). Each of these methods has its own advantages and disadvantages\, and the choice of which method to use depends on the specific requirements of the project. For our project\, we needed a low-cost\, accurate solutio n with the ability to efficiently perform millions of calculations. Curren tly\, no comparative analysis study evaluates or quantifies the existing m ethods for performing travel time calculations at the national level\, and there is no benchmark or guidance available for selecting the most approp riate method.\n\nTo address this gap in knowledge and choose the best driv e time estimator for our project we created a sample of 10\,000 ZIP/Hospit al pairs covering 49 of the 50 U.S. states with variable drive times rangi ng from a few minutes to over 4 hours. With this sample\, we calculated th e drive time using the Google Map API\, Bing Map API\, Esri Routing Web Se rvice\, ArcGIS Pro Desktop\, OSRM\, and OSmnx and performed a comparative analysis of the results.\n\nFor the Google\, Bing\, and Esri web services we used the Python requests package to submit requests and parse the resul ts. Within ArcGIS Pro\, we manually used the Route functions to calculate routes on a road network provided by Esri and stored locally. For OSMnx we utilized Python to perform the street network analysis using input data f rom OpenStreetMap. For OSRM we utilized C++ through the web API. OSRM prov ides a demo server to enable testing the routing without loading the road network data locally\, and we used this for calculating drive times for ou r 10\,000 samples. For generating visualizations we used Networkx and Igra h to display the shortest path of the drive time routing result\, and grap hs of our comparative analysis.\n\nWhen comparing drive time estimations u sing these 6 technologies we found: (1) There are very little differences among Google\, Bing\, OSRM\, ESRI web service\, and ArcGIS Pro when the ro ute drive time is less than roughly 50 minutes (2) For travel time estimat ions of routes greater than 50 minutes the Google and Esri methods were ex tremely close. The OSRM estimates produced travel times about 10% longer t han other methods\, and Bing’s estimates were about 10% lower than Googl e and ESRI. (3) Overall\, OSmnx estimates travel times lower than any othe r method because it estimates the shortest distance using the maximum velo city. In general\, the different methods employ different strategies for c onsidering traffic conditions. When long-distance travel is estimated the use of highways is required\, and each method employs specific parameters to account for traffic and resulting travel speed. Because of the complexi ty of modeling traffic conditions\, it is difficult to say which method pr ovides the most accurate and realistic driving times without empirical dat a being collected. Regarding cost\, the OSmnx and OSRM are both open-sourc e\, while the other methods have a cost for API usage (Google\, Esri\, Bin g) and desktop software (ArcGIS Pro). For processing efficiency\, Google\, Esri and Bing were all efficient\, each able to process the dataset in ro ughly one hour. We found the processing power of OSMnx was limited in the size of the road network it could handle\, so we had to divide the ZIP/Hos pital pairs into subsets by state\, and calculate them separately\, which was a laborious process. We found OSRM to be the most efficient\, able to handle 10\,000 requests in less than a minute. We ran OSRM in a high-perfo rmance cluster computing environment. This process included one hour of se tup to download the OpenStreetMap data for the entire U.S. onto the cluste r. Then we used Python requests to calculate the drive times and parse the result for analysis. The total processing time for the 32 million calcula tions ended up being 12 minutes.\n\nUsing OSRM provided us with a low-cost \, accurate\, efficient solution to calculating drive times between 32M or igin/destination pairs. We feel our study provides valuable guidance on ca lculating drive time in the United States\, offering a benchmark compariso n model between 6 different methods. We encourage others to utilize the co de produced for this project\; all of it is in the process of being publis hed on GitHub as open-source. Our analysis was just for the U.S.\, and per forming similar analyses in other countries will provide more insight into how useful the different methods are globally. In summary\, this comparat ive study allowed us to produce drive times in the most efficient manner i n order to support the larger objective of characterizing the quantity and quality of pediatric hospital capacity in the U.S. DTSTAMP:20260612T231836Z LOCATION:UBT E / N209 - Floor 3 SUMMARY:A Comparative Study of Methods for Drive Time Estimation on Big Geo spatial Data: A Case Study in the U.S. - Xiaokang Fu\, Devika Kakkar URL:https://talks.osgeo.org/foss4g-2023/talk/P7D3SC/ END:VEVENT END:VCALENDAR