Estimating corridor travel time mean and variance from its data sources.

Intelligent transportation systems (ITS) technologies and infrastructure are a rich travel time data source for estimating travel time mean and reliability. This paper describes how ITS data can be used to provide travel time information specific to commercial vehicles and to perform transportation system monitoring. Data were collected along a two-mile (3.2-kilometer) freeway segment in Houston (US 290) instrumented with automatic vehicle identification (AVI) at 0.5-mile spacing. Data were also collected along a two-mile (3.2-kilometer) freeway segment in San Antonio (IH-35) instrumented with inductance loop detectors at 0.5-mile spacing. Commercial vehicle travel time data were simultaneously collected by video along the corridor for comparison to AVI and inductance loop detectors. Instrumented test vehicle data were also simultaneously collected using distance-measuring instruments (DMIs) along the corridor at less than five-minute headways. The results of this study provide insight into how corridor travel time mean and variance may be estimated from both AVI and inductance loop detectors, and how these estimates compare to estimates of mean and variance for commercial vehicles. It was found that it may be reasonable to provide travel time maps and information in real-time specifically for commercial vehicles for just-in-time or fleet operations. ITS travel time data sources, such as AVI in Houston, were also found to provide a very cost effective data collection method. Further, the research demonstrates that test-vehicle runs should be periodically performed to evaluate the effectiveness and accuracy of the ITS infrastructure that is in place, and the DMI technology has been shown to be adequate for this purpose. Finally, the dependence of adjacent links is investigated by computing corridor travel time mean and variance for comparison to ITS data services. The information presented is anticipated to be useful for engineers and planners that use point or link ITS data for travel time estimation for real-time or off-line transportation applications.