Improving Supply Chain Logistics with Agent-Based Spatiotemporal Mechanistic Enumeration and Probe Vehicle Data

Abstract

Traditional methods of transportation performance and land use valuation rely on metrics that emphasize daily traffic volume and ideal travel speeds of transportation systems; however, enterprise logistics are better informed through data-driven models that identify reliable transportation routes between origin and destination. Enterprise and personal logistics are prone to disruptions from the variability of travel times across hours and days of the week. There is a critical need to assess and monitor the performance of global supply chains from the perspective of freight operations. In this work, a data-driven agent-based spatiotemporal mechanistic enumeration model was developed to evaluate the variability of completed round trips based on departure time, seasonality, and freight transaction times. The mechanistic enumeration methods utilize probe-vehicle travel time data that have been collected from devices equipped with Global Positioning System (GPS) receivers. Based on the success criteria of completed round trips, the results were evaluated to explore how operating conditions for an origin site (e.g., distribution center) and destination (e.g., maritime port) are influenced by the inherent variability of highway traffic performance and freight handling times. Supply chain logistics often focus on average highway travel times between an origin and destination to investigate freight round-trip duration and planning. Although the average drive time is easy to measure and communicate, there is no information on how daily and weekly traffic patterns will influence logistics. Using recent advances in GPS data collection and data processing, this study demonstrated how a computer-based mechanistic enumeration method is used to investigate freight operations performance. The methods provide a tailored approach that affords more information than traditional methods that explore average conditions. The mechanistic enumeration uses high-resolution historical traffic data to determine how the successful number of round trips is influenced by departure times, days of the week, months of the year, and transaction handling times (e.g., truck turnaround time). This information can be used by operators and truck drivers making decisions about schedules, hours of operation, days of service, and initial site selection, by considering how variable highway traffic conditions will influence logistics.