An interactive scheduler for production activity control

Traditionally, scheduling research has concentrated on the classical Tl machinclm job problem. Many of the algorililllis and scheduling systems developed can produce optimal or ncar-optimal solutions to well-defined scheduling problems. However, often there arc so many assumptions made in the process thnt, in the complex environment of the typical shop floor, the solutions proposed arc of little practical value. Their inability to lake into account real-life problems such as balch spliuiug, machine breakdown and maintenance, material availability CIC., means that unrealistic schedules often result. This paper recognizes these complexities, not by attempting 10 design a loudly automated scheduling system, which for the reasons described is impractical in most cases, but by including human interaction as an integral part of the solution strategy. Accordingly. the principles of socio-rechnical design have played all imporumt part in the development process. The result of this development. the interactive scheduler, is a decision support s)'stelll thai attempts to achieve a balance between available computing pO\\'er and human intuition and experience, the rome I' to alleviate the more mundane and tedious scheduling tasks and the latter to provide the intelligent input necessary to generate a solution. This solution will seldom be optimal, but it will usually model a given manufacturing system more closely than a schedule produced automatically, since the user has played an explicit part in its development.