Developing software for generating pouring schedules for steel foundries

Abstract

The authors have been involved in an industry-wide project whose goal is to improve the production systems of steel foundries. As a part of this effort, we have studied the process by which the foundries develop their daily pouring schedule. Typically, the foundries are unable to consider the impact of their pouring schedule on downstream operations because developing a feasible schedule is a labor-intensive task involving numerous constraints. To address this, the authors are working with Harrison Steel to develop software that can both automate the current scheduling process and improve it by estimating the impact of the pouring schedule on downstream work-in-process (WIP) inventory levels. An integer programming model is described that minimizes a comprehensive cost function that includes the costs of pattern tooling set-up, late delivery, WIP inventory, and under-utilization of assets. The software implements a heuristic that finds multiple solutions to this integer program, each of which corresponds to a feasible schedule. Computational tests reveal that the software is capable of handling realistically sized scheduling problems in a reasonable amount of time. Upon its completion, this system could be extended for implementation at many other steel foundries.