From Microsimulation to Nanosimulation

Abstract

In the past 15 years microsimulation software tools have increased the ability to analyze congested traffic by modeling at the level of individual vehicles. Although this can be very detailed, it assumes that mode choice is fixed. The research presented models the people in the network, either walking or in vehicles, following each person for an entire trip through multiple modes of travel. This approach is called “nanosimulation.” The paper presents a pilot project that analyzes access to an airport by comparing multiple parking options, rail transit, drop-off, and taxi access. Generalized cost incorporating time, distance, and price is visualized for each access method and allows comparison of the total end-to-end cost of all combinations of modes in an interactive three-dimensional simulation model. The primary objective of the research presented is to prove that analysis at this level is practical and can provide insight that is not available from other methods. A historical perspective of microsimulation is presented to illustrate how seemingly impractical models that were run on a supercomputer in 1994 are now in common use on sub-$1,000 everyday computers. The paper describes the technologies, data structure, and algorithms used in the research and addresses the issues of data availability and model repeatability. The model area is described, with illustrations of the visualization. The final sections present results of the project, followed by conclusions and recommendations.