A framework for artificial transportation systems: from computer simulations to computational experiments

Abstract

This paper outlines the basic concepts, major methods, and key issues for the development of artificial transportation systems, which is a natural extension of computer simulations for systems analysis and decision support of transportation systems. The main idea is to integrate various transportation models into "artificial" transportation systems and convert computers into experimental "fields" for transportation analysis and decision making and evaluations. The keys to successful applications of such artificial transportations are the availability of agent-based programming and modeling, large scale distributed computing techniques, and new concepts and methods developed in complex systems, such as, artificial societies, computational experiments, and parallel systems.