The San Diego Association of Governments (SANDAG) is enhancing its capabilities to analyze projects that improve system efficiency over capacity expansion. Measuring the benefits for such projects with traditional static equilibrium traffic assignment tools can be challenging for a number of reasons; their lack of key transport network features such as traffic signal timing and their lack of queuing being most important and often-cited among them. Therefore SANDAG has begun the development of a fully-integrated Dynamic Traffic Assignment and Activity-Based model. The Aimsun dynamic traffic assignment software package and the existing SANDAG Coordinated Travel-Regional Activity-Based Modeling Platform (CT-RAMP) is being utilized for this project, which will involve significant algorithmic and software enhancements to both systems. The resulting modeling system will greatly improve SANDAG’s ability to measure the demand for and benefits of managed lanes, toll facilities, intersection improvements, signal timing, ramp metering, incident response mechanisms and other similar traffic management and operations solutions.

A key facet of this work is the use of a consistent Geographic Information System (GIS) database for the management of detailed network data required by the Aimsun DTA package and for other agency GIS support needs. This paper describes the development of the GIS database schema, the collection and coding of updated network data including intersection geometry within the Aimsun DTA model, the storage of all network data within one GIS database, and the process used to import data stored in auxiliary databases including the San Diego Regional Arterial Management System (RAMS), Ramp Meter Information System (RMIS) and Congestion Pricing System (CPS). The paper describes the fully-automated process developed to import all external data and create base and future-year networks from that data and GIS layers. The paper further describes enhancements made to the CT-RAMP model and aggregate commercial vehicle and other models to prepare disaggregate trip lists for initial network validation; converting from person trips to drivers, temporal and spatial disaggregation . The key challenges that the project team has faced and solutions to those problems are described.