Repetitiveness of daily travel

This paper describes research that explores the use of habitual and repetitive patterns of travel behavior for improving travel demand modeling. The data used in this research came from two Australian panels from which Global Positioning System (GPS) data were collected. The trips recorded by GPS were initially classified into tours. A tour classification that grouped the tours into 12 tour types was then applied, and the number of repetitions of each tour type within the time period of data collection was determined. Having determined how often a particular tour type was repeated over a week or longer, the authors then described tours by certain tour characteristics as the initial means to search for identical patterns of travel from one day to the next. To determine whether a particular tour type was repeated more or less identically on more than one occasion, the tour type repetitions were described by a coefficient of variation for each of tour travel time, tour distance, tour activity time, and total duration (time) of the tour. From this process it was determined that relatively little repetition of tours occurred from one day to the next; this finding raises serious questions about the assumptions of repetitiveness that underlay almost all travel demand models.