ARE WE THERE YET? USING GPS TO TRACK DRIVE TIMES IN DAVIS, CA

Title Are We There Yet? Using GPS to Track Drive Times in Davis, CA.
Author David M. Soule American River College, Spring 2006 Geography 350: Data Acquisition in GIS
Abstract Concerned with apparently excessive driving time while running daily errands in Davis, Ca, this novice GPS user tracked his drive times around town for two months for the purpose of formulating a database of drive time data for eventual comparison to City of Davis data
Introduction This is largely a data acquisition project established for the purpose of collecting as much data as possible in order to compare driving times across town on various routes and to compare real world results with drive times provided by the City. It required much labor intensive data processing to get the data into a state to make it comparable to the City data. But the results show it is both feasible and worthwhile to use a GPS to study drive times in Davis.
Background Davis is known as a bike friendly town. But it has grown nearly to its limits and yet continues to bring in new people. More students are driving across town rather than riding bicycles as it had been famous for and the University has constructed more than one new multi-level parking garage to accommodate them. More stop lights, with cameras and signs announcing formidable fines for running red lights, are going up throughout town. The east-west layout of the town limits the routes one can choose to take across town. Finally, the University is proposing the construction of housing for at least another 4-5,000 people on now vacant university land immediately west of Highway 113. All of this points to increased drive times for those unwilling or unable to bike across town.
Methods There were three steps to this project: data acquisition, data processing, data comparison and initial analysis. Data acquistion was performed using a Garmin GPS76. The data was acquired "randomly" in that no schedule was made to collect it, and no special routes were planned. Rather the data was collected "ad hoc" -- in the normal course of conducting daily business. Such business includes trips to school, church, the grocery store, the post office and the credit union. The east-west layout of the town and my location on the western edge further actually serves to limit the available routes to go across town. I collected data for about two months. Upon filling the GPS track log I downloaded it into DNR Garmin. Initially I noted the information collected on the trip computer as I arrived at each destination. This often became onerous as a result of such things as lighting, (at night the screen was difficult to read, especially without glasses) my occasional lack of time, or even lack of discipline. Later the GPS stopped recording "stopped time" during the tracking. Eventually, when downloading the data into DNR Garmin I would take screen shots of the "track" screen, which gave total distance for the track, and of the saved map. The data acquistion phase included accessing the City of Davis web site to make use of the many layers available there, the most important being streets, stop lights and city limits. The streets layer has attribution for street segment length and travel time. The next step in this project was to process my data to make it measurable and comparable. First I analyzed the various tracks and made new layers every time a TRUE value was given in "new segment" attribute for Active logs. It was necessary to use the Active Log type rather than the Track point type because the Track points do not give a unique time of acquisition. For every point (feature) the Active Log time attribute is the UTC for the actual moment the point was laid down. These then are truly unique features. Doing this provided me with a set of complete and discrete track that modeled my individual trips and allowed me to count elapsed time. The processed layers I then placed down on the Davis street segments in ARCGis and readied them for comparison to the City of Davis data. Relative to the courses I traveled and the City data regarding the streets on which I traveled the GPS device places track points "randomly". In other words a track point is not necessarily placed exactly at the beginning of my course, or when I turn a corner. This part of the data acquisition process I had essentially no exact control over. The City of Davis, however, measures its streets in very small and precise segments. To prevent the possible comparison of apples and oranges I had to choose end points on my tracks that, as best as possible, coincided with City street segment endpoints. Although this was not hard to do, it does insure that there is some error in the comparison. At this point The error would be negligible because Davis street segments are so small that virtually every segment was significantly less than 60 seconds. At this point the work became somewhat laborious as I selected the numerous Davis street segments, short as they were, under my track, saved as a layer, and then summarized the "seconds" attribute. This gave me the amount of time necessary, according to the City of Davis, to travel the route I had taken if I had been travelling according to their maximum or average speeds limits and/or average speeds. I then turned to my track layer, identified my two track endpoints that most nearly coincided with the city segments and calculated the elapsed time. With these two numbers I can then make a meaningful comparison of my actual drive time and the City assumptions about what it takes to get from point to point in rive time and compare it to City of Davis.
Analysis Initial analysis and comparison of my data showed that it takes much longer to get across town than is allowed for in the City data. For example the trip from my house to my son’s elementary school should have only taken 5 minutes and 12 seconds. It actually took me 17 minutes and 21 seconds. A trip from that school to the post office and then home, not including time, waiting to pull out of the post office lot – which is a difficult choke point in Davis, should have taken 13 minutes according to the City of Davis. It actually took me 20 minutes and 11 seconds.
Conclusions My data shows that there may be something to the complaint, near universal in Davis, that things aren’t what they use to be or what they seem to be. The issue certainly bears further investigation and further development of the data already acquired. Further investigation should probably include interviews with the City of Davis to discover how they formulated their data regarding drive times. It may be that I was in fact, comparing apples and oranges. Or it may just be that the truth, my data, is stranger than fiction, the City data.
References City of Davis GIS Layers http://www.city.davis.ca.us/gis/library/details.cfm?FileID=31 How Do Students Get Around in Davis? A Third Year Davis Honors Challenge Seminar March 14, 2002. Athena Chuang, Crystal Castaneda, Patricia Forman, Maria Nozzolino and Mariah Talso