Practice of Management Science---Orderbook Balancing Using a Combination of Linear Programming and Heuristic Techniques

In complex manufacturing situations, when product demand exceeds the manufacturing capacity by a very wide margin, both the production planning and scheduling functions become extremely difficult and time-consuming to handle---even with modern computers and sophisticated analytical techniques. Our studies had shown that in such cases the tasks of production scheduling and detailed planning could be made easier if the orderbook were first balanced in an aggregate manner such that monthly or other periodic demands were approximately in line with the manufacturing capabilities.We applied this basic approach in the development of an improved planning tool for Bethlehem Steel's roll manufacturing operations, where approximately 4,000--6,000 cast rolls of 300 different types, each requiring an average of ten operations, are machined in a large m / n job shop each year. This planning tool, which is a hybrid computer package combining linear programming and heuristic methods, is being routinely used to balance the roll orderbook. The quick estimate of shop capacity and due-date performance provided by this package gives shop management an improved ability to negotiate roll delivery and resolve order-distribution problems with the plants. Due dates are changed and/or orders are dropped during one or more rounds of negotiations with each plant, and a balanced, or feasible, orderbook for the plants as a whole is obtained. Accurate production schedules and long-term plans are then generated by elaborate simulation models.