The reliability history of airtrans

The totally automated AIRTRANS system, bought by the Dallas/Fort Worth Airport Board and designed and built by the Vought Corporation, Dallas, Texas went into revenue service on the day the giant new airport opened in January, 1974. Since that day, it has served as a generally convenient and rapid connector between the airport's widely spaced passenger and employee stations. Over 13 miles of concrete guideway, 51 totally automatic and driverless passenger vehicles and four automatic utility vehicles circulate 24 hours a day over several loops and through 14 passenger and 13 employee stations. The system is controlled by a central computer and several regional computers. At a maximum speed of 17 mph, the longest trip time for any passenger between any two terminals, including wait time, is 20 minutes (30 minutes from a terminal to a remote parking lot). Most trips are much shorter - except when equipment failure occurs. Records kept by the central control operators - currently two persons per shift, 24 hours per day - have indicated generally a high level of availability after the initial system bugs were eliminated. In one good period the system performed for almost 200 days without a failure-induced delay long enough to justify calling out the backup busses. To permit other automated system specifiers and designers to benefit from the lessons learned in the AIRTRANS experience, the Urban Mass Transportation Administration has sponsored two assessments of the system at different stages of its life. These were performed by the Transportation Systems Center, and the most recent of them collected and analyzed the reliability, maintainability, availability and safety information on the system as it was recorded in the central operators logs over a period of three years. A sampling of the data was done for each of the six periods of the system's life, which were arbitrarily defined as being periods of operation during which the stable condition of the system differed from that of previous periods. Thus the first period of 147 days saw passenger service for 15 hours per day; the second of 217 days saw passenger service for 24 hours per day; the third period of 56 days saw both passenger and utility service for 24 hours per day; during the fourth period of 182 days Vought maintained the system on contract to the Airport Board. During the fifth period of 70 days the Airport Board took over maintenance itself. The sixth and final period of 238 days saw the system expand to include an employee transport service, using dedicated vehicles. An adequately large sample was taken to give good confidence that the various indices of performance system reliability, availability, mean miles between system failures, mean time to restore the system, and many others - were representative and meaningful for each period. Dennis Elliot of the Dallas/Fort Worth Airport Board extracted the raw data from the logs and maintenance records in accordance with a plan developed by TSC; and the data were reduced by TSC. The complete results will be contained in an AIRTRANS, Phase II Assessment Report, soon to be published by TSC. This paper is a preliminary Summary of the data taken, analsis done, and the performance, both of the system and its subsystems, as revealed by the analysis. This is real life data taken on an operating automated guideway transit (AGT) system. It is hoped that it will be of value to system planners and to designers both of systems and of their component hardware elements.