TRB 2016 Blue Ribbon Committee
16th National Transportation Planning Applications Conference

Visual representations of regional mobility with the MAG ABM


Corresponding Author: Daniel Florian, INRO

Presented By: Daniel Florian, INRO

Coauthors:

  • Vladimir Livshits, Maricopa Association of Governments
  • Peter Vovsha, WSP
  • Kyunghwi Jeon, Maricopa Association of Governments
  • Petya Maneva, Maricopa Association of Governments

Abstract

The Maricopa Association of Government's recently launched activity-based travel demand model is an example of the trend towards demand microsimulation and agent-based modeling, and like other 'new-generation' travel demand models it is characterized by a disaggregate representation of activities and mobility performed by a fully realized synthetic population. For example, a base year 2015 run of the new MAG activity-based travel demand model produces approximately 20 million activities and 16 million trips that take place at a temporal resolution of 1 minute and a spatial resolution of twenty-six thousand micro-analysis zones.

In terms of data scale and complexity, the boundary between observed Big Data and simulation output becomes blurred, and new visualization and analytical tools are required to understand the data. This prompted a review for new methods of analysis at MAG starting in 2015. Recently, MAG, WSP|PB and INRO developed a visualization and analytics interface for the new MAG ABM with promising results. The interface transforms the ABM model outputs into a visual representation of urban mobility in the region that can be used to interactively visualize and analyze model results. Visualizations were used for communication purposes to executives and main stakeholders and helped to generate interest in a variety of ABM applications, but were also used effectively to interpret model results for quality control and validation, trend identification and scenario development. The presentation will provide application examples in the MAG region.

Presentation

Paper

Discuss This Abstract