Computerized schedule construction for an airbus transportation system

Abstract

As part of a continuing study of the application of V/STOL aircraft to the transportation problems of the Northeast Corridor, the U.S. Department of Commerce has requested that detailed data be developed on schedules, travel times, and fares which might be expected for a V/STOL system operating in the year 1980. This section deals with the computer methods used to construct such schedules. A schedule (or more properly a schedule plan) is a complete description of a transportation system. It details the services to be offered in the dimensions of time and geography, gives the routings followed by vehicles, and indicates the loadings to be placed upon terminals. A complete statistical summary of the operation of the transportation system can be obtained once the schedule is completed. The number of vehicles and crews, their daily utilization, the expected load factors, the required number of loading gates, the average length of vehicle hop, etc. are all implicitly determined by the schedule plan. Constructing and maintaining an efficient schedule is the main problem of transportation system managements. It is both their production plan and their product to be marketed, and the economic success of the plan is gauged by the management's ability to produce a low cost production which will be saleable to the travelling public. The use of computers in scheduling is not widespread at this time, and if they are used, it is generally for data processing as distinct from decision making or problem solving. The reasons for this are clear. There has not been in the past, sufficient capability either in the hardware, or the software to handle problems of the size and complexity associated with even such relatively small transportation systems.as the airline systems. This situation has been changing in the last few years, to the point where we can now begin to handle fairly large scale scheduling problems, introducing optimization at several points, and constructing fairly quickly and easily full system schedules and their statistical summaries. Parametric investigations of the effects of restricting fleet size, terminal size, etc. can be quickly carried out. Various strategies or policy decisions are similarly easily investigated. The construction of computer programs for the scheduling process immediately points out the need for detailed accurate data concerning demand. This is now becoming available to the airlines through their reservation systems and the management information systems evolving from them. We need to know, for example, detailed information on the number of people travelling from A to B throughout the day, by day of week and month of year, with an accuracy much greater than our present estimates. We would like to know the demand elasticity, e.g. the change in the number of people travelling as services are changed in time or quality for every service pair in the system. The character of the available data (as distinct from opinion) determines the type of problem that operations research and the computer will be able to successfully solve. Various large scale econometric models are conceivable, if revenue and cost data are available. This section will describe the work which has been carried out for a hypothetical Airbus short haul V/STOL system in the Northeast Corridor. It is only a beginning as valuable extensions are yet to come.