Cross-cutting Issues

Developing a frontal collision warning system (FCWS) for transit buses is considerably much more difficult than that targeted to highway applications due to the significant differences in operating environments. In this paper, a histogram analysis was applied to several months of real-world data collected on an experimental bus in normal service in San Francisco Bay Area. This analysis addresses the unique characteristics of the operational environment of transit buses. Figures and pie charts as well as analyses are presented.

University of California, Berkeley

Presented at the ITS America Annual Conference and Exposition, May 19-22, 2003 Minneapolis, Minnesota

The City of Palo Alto, which owns and operates 96 traffic signals, is in the process of implementing an Advanced Transportation Management System (ATMS) that will include features such as adaptive traffic control and integration with ITS devices. This spring the City will select a vendor to implement an ATMS comprised of a central signal system server located in their traffic management center (TMC), 25 non-actuated NEMA controllers for their downtown intersections and 71 Model 2070L controllers for the rest of their network. All of the 2070L controllers will be equipped with Ethernet ports and Ethernet switches and will be connected over the City’s fiber optic network to an Ethernet switch in the TMC. The Central server will communicate with all of the controllers using Ethernet. The system will be compliant with National Transportation Communications for ITS Protocol (NTCIP) standards.

Kimley-Horn and Associates

City of Palo Alto

Presented at the ITS America Annual Conference and Exposition, May 19-22, 2003 Minneapolis, Minnesota

An Intelligent Transportation Systems (ITS) communications architecture is emerging with ITS-specific standards for several layers of the International Standards Organization Open Systems Interconnections (OSI) communications reference model. However, these standards neither fully leverage the advantages of a standards-based architecture, nor fully address ITS-specific security and networking requirements. This paper rectifies these shortcomings by describing a reorganization of the emerging standards and the addition of protocols for two additional OSI layers. The result is an improved communications architecture for ITS networks.

The George Washington University

Presented at the ITS America Annual Conference and Exposition, May 19-22, 2003 Minneapolis, Minnesota

Data from inductive loop detectors is widely used as input to many ATMS applications. However, for several reasons, loop detectors are prone to failures. This leads to erroneous data that significantly reduces the accuracy of the applications that utilize detector data as input. The work that has been done in the past to develop procedures to test the quality of detector data is inadequate since there are several forms of corrupt data that still remain undetected. Moreover, procedures to correct erroneous data identified are also lacking. This paper presents an improved procedure to test the reliability of detector data as well as a procedure for correcting erroneous data. The data screening and correction procedure is applied to detector data from sensors of a real freeway section to demonstrate its effectiveness and applicability.

University of Minnesota

Presented at the ITS America Annual Conference and Exposition, May 19-22, 2003 Minneapolis, Minnesota

A preliminary study revealed that Washington State Department of Transportation’s dual-loop detection system was not functioning properly. In order to identify types and severity of dual-loop errors, an in-depth evaluation on the accuracy of aggregated dual-loop data was performed. The sample dual loops were all under optimal working condition, which minimized the possibility that system failure resulted in inaccurate loop measurements. Aggregated dual-loop data were compared to video data on 20-second intervals so that errors were captured whenever they occurred with an error range of ±20 seconds. Two major errors, miscount and misclassification, were quantitatively identified; possible causes of dual-loop errors were investigated; further research directions were also recommended.

University of Washington

Presented at the ITS America Annual Conference and Exposition, May 19-22, 2003 Minneapolis, Minnesota