Selecting and Modeling the Phase 1 Test Area

Based on the information collected about the corridor, areas with good data availability were identified. One such area, in Arcadia along the I-210 corridor between Michillinda and Huntington Dr., was selected as the Phase 1 test area:

This Phase 1 test area was then modeled and simulated, providing a venue to assess each step of the AMS methodology.

  • Freeway: The model includes about 18.5 miles of the I-210 freeway in the westbound direction.
  • Arterial: In addition, it includes a parallel route along Huntington and Colorado through Arcadia and part of Monrovia. The arterial is modeled in both eastbound and westbound directions, including all signalized intersections and their cross streets.

The overall workflow for building and calibrating the model looks like this:

  1. Analyzing and cleaning the data—The first step of the workflow is to clean and aggregate the various types of data collected (e.g., loop detector data, traffic study data, signal timing sheets).
  2. Calibrating the model components—These data are then used to calibrate the CTM model to reflect observed traffic conditions, which means to decide the best values to use for parameters such as fundamental diagrams and inputs such as boundary flows and split ratios.
  3. Running the model—The CTM simulation model is then run, which produces measures of density, velocity, and flow.
  4. Analyzing/aggregating the results into metrics—Performance metrics like travel time, delay, level of service, vehicle hours traveled, and vehicle miles traveled are calculated from the simulation results.
  5. Validating the system against reality—The simulated travel times and flows are compared against those measured in the field. If there are significant discrepancies between simulation and field measurements, the calibration process is repeated until results pass a reasonability check.

Working through this process generated confidence in model choices, algorithm performance, and scalability. For complete details on model construction, calibration, and validation in AMS Phase 1, see the AMS Phase 1 Report.