FDOT Reliability Performance Measures

KAI, along with FDOT's Central Office and the University of Florida (UF), developed models for predicting travel time reliability for freeways, using data from Philadelphia and Ft. Lauderdale. These tools can quantify travel time reliability as a function of various changes in the system, such as incident removal times and work zone occurrences, as well as selected ITS programs and initiatives (such as the Road Rangers). These procedures were implemented to estimate travel time reliability on the entire freeway portion of the SIS, for multiple years. The reliability of travel time for various trips can be used in three types of planning activities: to assist in policy decisions, for system-wide project planning, and for project development activities. This reliability model is based on a level-of-service (LOS) database, provided by each District by location on an annual basis.

Using this database, KAI and UF developed a model to convert average annual daily traffic (AADT) volumes to hourly directional traffic volumes in order to estimate hourly directional “base” speeds. Beginning with these “base” speeds, the following input variables (with a probability of occurrence) are used individually and in combination with the others to calculate a revised speed for each hour:

  • Probability of demand over capacity
  • Probability of rain
  • Average precipitation by hour
  • Average number of closed lanes per incident
  • Probability of work zone
  • Average number of lanes closed per work zone

Using the result of the combination of scenarios noted above, travel time reliability, based on on-time arrival, buffer index, and travel time index, were calculated.

In Phase 3 work, KAI assisted UF and FDOT with research regarding the implementation of Travel Time Reliability into FDOT's annual reporting program. This project was a part of KAI's contract, and the objectives of this project were to: 1) identify likely performance measures to be reported by this tool (for annual reporting to the legislature, for project prioritization, for interchange approvals, etc.); 2) develop new models and expanded methods for obtaining travel time reliability measures for arterials and multimodal corridors; 3) develop a framework for expanding the travel time reliability estimation methods to all modes of travel (rail, transit, water, and aviation); and 4) modify the existing database tool to incorporate the new research results.