On the formulation of time-space prisms to model constraints on personal activity-travel engagement

The notion of time-space prisms has often been used in the context of descr ibing activity-travel patterns of individuals. This paper presents a method ology for estimating the temporal vertices of time-space prisms using the s tochastic frontier modeling technique. Observed trip starting and ending ti mes are used as dependent variables and socio-economic characteristics and commute characteristics serve as independent variables. The models are foun d to offer plausible results indicating that temporal vertices of time-spac e prisms, though unobservable, can be estimated based on temporal character istics of observed activity-travel patterns. Comparisons of stochastic fron tier models of prism vertices and the distributions of prism vertices are p resented using two activity data sets collected in the United States - San Francisco and Miami. Differences and similarities in temporal vertex locati ons are highlighted in the paper.