MEDIUM TERM PRODUCTION PLANNING AND CAMPAIGN SCHEDULING FOR SAWMILL

Abstract

In this thesis, we study a multi-period, multi-product, production planning problem for the lumber industry, and present a hierarchical approach to control and schedule lumber production in a sawmill. First, at the sawing unit, the lower level of the hierarchical structure, the combination of log classes, price lists and sawing patterns defines the expected output distribution in terms of lumber pieces. A price list, defining the value of outputs, is fed directly to the sawmill production control optimizer to select the best sawing patterns. This results in a broad variety of different lumber outputs. Second, at the upper level, a mixed integer programming model has been proposed to maximize the total revenue at the sawmill. The lumber outputs determined at the lower level are used as data at the upper level. Market demand, lumber inventory cost, and supply cost are considered over the planning horizon. The proposed model has been developed and implemented on a real-scale prototype sawmill.