A THREE-STAGE CONTROL MECHANISM FOR THE LUMBER PRODUCTION PROCESS OF A SAWMILL BASED ON A POWERS-OF-TWO MODELLING APPROACH

Abstract

To control the lumber production in a sawmill, a three-stage system is proposed. First, a quick program creates many cutting patterns and chooses the most valuable pattern for each log within a log class given a price list. A combination of a log class and a price list resulting in a set of lumber output proportions creates a “campaign”. Second, a Powers-of-Two optimization model calculates “campaign lot sizes” to minimize inventory and meet deterministic demand. The goal is to develop a minimum cycle stock inventory level for all the products over a time horizon. Third, five control approaches are created based on the results of the PoT model and evaluated using simulation environment to monitor inventory levels in the case of stochastic demand. This research indicates that in a divergent-stochastic environment such as a sawmill, situations with noisy batch order arrivals do pose difficulties when Powers- of-Two control approaches are used.