Production scheduling and control have long been the subjects of extensive research. In many cases, production scheduling becomes so complex that ill-advised decisions are made, resulting in higher costs and longer delivery times. The purpose of this work is to build, document, and validate, in real factories of three types, a usable, useful software system to cope with these complexities. The approach entails three phases: production planning in multi-stage ("flexible machining") systems, incorporating routing choice; planning for uncertainty; and short-term scheduling. The outputs will include models which provide a master production plan (typically deterministic); models which determine the sizing and location of safety stock in the production system (to provide against stochastic changes in demand, supply, or yield); and provision for linkage of the models into a system. The work will be tested and evaluated by simulation, using computer graphics to display Gantt charts, inventory-time plots, spatial flows of materials, and monitoring of variables.
