Conversion of Volunteer-Collected GPS Diary Data into Travel Time Performance Measures: Algorithm for Extracting Travel Diary Data from GPS Streams and GPS-TDG Software Design

Conventional travel-survey methodologies require the collection of detailed activity-travel information, which imposes a significant burden on respondents, thereby adversely impacting the quality and quantity of data obtained. Advances in global positioning system (GPS) technology have provided transportation planners with an alternative and powerful tool for more accurate travel-data collection with minimal user burden. The data recorded by GPS devices, however, do not directly yield travel information; the navigational streams recorded by GPS devices have to be processed and the travel patterns derived from them. The focus of this research project is to develop software to automate the processing of raw GPS data and to generate outputs of activity-travel patterns in the conventional travel-diary format. The software will identify trips and characterize them by several attributes, including trip-end locations, trip purpose, time of day, distance, and speed. Within the overall focus of the research, this report describes the algorithm developed for the detection and characterization of trips from GPS navigational streams. The software design details are also presented.