Search

Vehicle classification data are used in many transportation applications, including: pavement design, environmental impact studies, traffic control, and traffic safety. Ohio has over 200 permanent count stations, supplemented by many more short-term count locations. Due to the high costs involved, the density of monitoring stations is still very low given the lane miles that are covered. This study leveraged the deployed detectors in the Columbus Metropolitan Freeway Management System (CMFMS) to collect and analyze classification data from critical freeways where the Traffic Monitoring Section has not been able to collect much classification data in the past due to site limitations. The CMFMS was deployed in an unconventional manner because it included an extensive fiber optic network, frontloading most of the communications costs, and rather than aggregating the data in the field, the detector stations sent all of the individual per-vehicle actuations (i.e., PVR data) to the traffic management center (TMC). The PVR data include the turn-on and turn-off time for every actuation at each detector at the given station. Our group has collected and archived all of the PVR data from the CMFMS for roughly a decade. The PVR data allows us to reprocess the original actuations retroactively. As described in this report, the research undertook extensive diagnostics and cleaning to extract the vehicle classification data from detectors originally deployed for traffic operations. The work yielded length based vehicle classification data from roughly 40 bi-directional miles of urban freeways in Columbus, Ohio over a continuous monitoring period of up to 10 years. The facilities span I-70, I-71, I-270, I-670, and SR-315, including the heavily congested inner-belt. Prior to this study, these facilities previously had either gone completely unmonitored or were only subject to infrequent, short-term counts.

Traffic data is used to estimate current traffic conditions so that travelers and agencies can make better decisions about how to use and manage the transportation network. This research explores the fusion of probe data (vehicle speed and direction) with loop data (density, speed, and count) in the context of producing overall network speed and travel time estimates. Speed and travel time estimates are useful in many circumstances, but current system control strategies (ramp metering, for example) require density data. While it is difficult to significantly increase the quantity of loop detectors on state highways, the penetration rate of probe data is continually increasing. Multiple data sources with various characteristics were fused by running probe and loop data through the Mobile Millennium highway model, generating velocity maps and travel times. The performance of data sources both individually and when fused was evaluated. It was found that the highest quality estimates are achieved by combining probe data and loop detector data.

One of the most common traffic volume parameters reported by statewide traffic monitoring programs is annual average daily traffic (AADT). Departments of Transportation (DOT) and other state agencies use a series of continuous vehicle detection devices in association with smaller more mobile short-term counts. Once the short-term counts are recorded a series of adjustment factors (time of day, day of week, month of year, or seasonal) are applied to the short-term counts. The end result is an estimated AADT for a particular segment of roadway. Traditionally, as defined in section two of the Traffic Monitoring Guide (TMG), there are three methodologies, geographic/functional assignment of roads to groups, cluster analysis and the same road application factor. In each case, there are advantages and disadvantages and currently there is not a final peer reviewed nationally suggested method. The benefits associated with this research include an improved method for estimating AADT throughout Ohio.

The purpose of this guide is to aid the Texas Department of Transportation (TxDOT), Metropolitan Planning Organizations (MPO), and other state and local agencies to develop an effective traffic monitoring system for new major traffic generators in their metropolitan region. This guidebook will provide you, the transportation planner and engineer, with a better understanding of the considerations that need to be taken when developing a monitoring system for these types of large generators.

As the technological complexities of and public demands upon our Intelligent Transportation Systems (ITS) infrastructure increase, new opportunities and requirements arise regarding how best to manage existing ITS assets and select future deployments. This research project aims to support such decision making by developing methods that clearly relate sensor coverage (and other ITS data sources) to Dynamic Message Sign (DMS) performance via algorithms that predict freeway traffic time. The report documents the research conducted as a part of this project. A detailed literature review of the state-ofthe- art and the state of the practice in travel time prediction has been conducted and some of the limitations of the existing models have been identified. Two innovative travel time prediction models have been proposed and their performance has been tested with both simulation and real data. Both of these models use as input traffic counts obtainable from single loop detectors, which are the most widely deployed traffic sensor. The first model is based on an integrated statistical simulation framework and the second model is based on comparing cumulative counts. Methodologies have been developed to arrive at optimal detector location and spacing for better travel time prediction. Some of the issues involved in real-time deployment have been studied and summarized.