Citation

Dudek CL, Messer CJ, Dutt AK. Transp. Res. Rec. 1974; 495: 35-43.

Copyright

(Copyright © 1974, Transportation Research Board, National Research Council, National Academy of Sciences USA, Publisher SAGE Publishing)

DOI

unavailable

PMID

unavailable

Abstract

An experimental warning system has been installed on the Gulf Freeway in Houston as a means of alerting drivers approaching crest type of vertical curves to stoppages downstream of the crest. Successful automatic operation of the warning system depends on the reliability of system components. Earlier studies showed that developed control logic is responsive to stoppage waves, provided that the hardware functions properly. A one-lane control criterion resulted in 100 percent detection whereas 96 percent of the waves were detected using a two-lane control criterion. The studies also indicated a relatively high frequency of detector failures. The frequency of detector failures prompted a study to evaluate reliability of the warning system based on detector failure and repair rates experienced on the Gulf Freeway surveillance and control system and to ascertain whether detector redundancy or improved maintenance would be necessary. Malfunctions and repairs of all the Gulf Freeway surveillance and control hardware, including the 96-detector subsystem, were recorded for a 5-month period. The data revealed that the detectors on the Gulf Freeway failed at a rate of 0.000378 failures per detector hour. Detectors were repaired at a rate of 0.23 repairs per hour. The reliability in terms of availability of the safety warning system was analyzed using these data and classical models for maintained systems. Availability of the system was 0.95 and 0.995 for the one-and two-lane criteria respectively. The results indicated that the current detector configuration and maintenance practices were adequate.