Identifying and modeling urban truck daily tour-chaining patterns

Abstract

The main goal of this research is to better understand truck tour patterns in an urban setting and develop models that can describe daily tour-chaining patterns. This research uses truck activity data collected for the Urban Freight Heavy Vehicle Study ongoing in Singapore, which is an advancement in freight data collection studies. The data contain individual truck's Global Positioning System (GPS) traces and rich behavioral details including the activities at stops and operator's characteristics that were processed and verified though a freight data collection platform. Based on the initiative of using post-processed GPS data for tour identification, this paper refines the definition of tour and tour chain to explicitly reflect stop purpose, stop duration, and time of stop. Tour types and daily tour-chaining patterns in the dataset are identified. Further, this paper presents discrete choice models developed to explore factors that influence daily tour-chaining patterns. Identified important factors are: the difference between the number of distinct pickup and delivery locations, geographical spread of distinct pickup and delivery locations, shipment type, time to start work, employment type, land use type, and truck type. The major contributions of the paper are: 1) identifying limitations of the conventional definitions of tour and tour chain and proposing new approaches to reflect logistics practices; 2) explaining the tour-chaining patterns of heavy goods trucks in Singapore; 3) developing tour-chaining pattern choice models that aims serving agent-based simulation platforms.