Abstracts Business

An Agent-based Freight Business and Logistics Decisions Simulation - FREBUILDS

by Usman Ahmed

This dissertation focuses on agent-based microsimulation urban freight transportation modelling. Urban freight models are classified based on multiple dimensions of freight transportation. A conceptual framework of an agent-based urban freight model is developed. The framework integrates key business and logistics decisions, from short- to long-term, made by firms. Some of the modelling components of the framework are estimated for the Greater Toronto and Hamilton Area. This dissertation contains four empirical studies that are part of the conceptual framework of the urban freight model. The first study focuses on long-term business decisions and estimates an establishment location choice model that incorporates intra-firm interactions. The next three studies focus on key logistics decisions that are part of the shipment formation decisions. The first of these studies estimates models of freight transportation outsourcing and vehicle type choice decisions. The two decisions are modelled independently and jointly using discrete choice methods. The second study compares machine learning and discrete choice methods for the choice of freight vehicle type. The final study, that is part of shipment formation decisions, estimates empirical models of freight vehicle type and shipment size choice. It studies the nature of the choice process i.e., sequential or joint, and recommends the appropriate modelling structure. Together with the remaining modelling components of the urban freight modelling framework, this dissertation is a first step towards operational urban freight models.