Application of Metaheuristics to Packing Formation Support Systems of Pre-Cut Lumber Factory

Abstract

In recent years, wood has been considered as a building material from the perspective of SDGs. Pre-cut lumber is often used for wooden houses, including roofs, floors, and pillars. It is preprocessed into a certain shape for the joint parts and has joint brackets if necessary. The use of them has the advantages of shortening the construction period and reducing the construction cost. The number of pre-cut lumber production factories in Japan decreased from 757 in 2001 to 659 in 2011 and then increased to 730 in 2016. Compared with 2011, the number of factories with sales of <500 million yen in 2016 decreased by approximately 30%, while the number of factors with sales of ≥500 million yen increased by 80%, indicating a trend toward a larger scale [1]. Therefore, improving the productivity of these factories is imperative. We have been researching the conversion of factories into smart factories to improve the productivity of pre-cut lumber manufacturing company A. In this company, employees take 1.2 h to prepare packing plans every day, and the company faces the challenge of improving the efficiency of planning operations. Herein, we propose an algorithm using an iterative local search (ILS) with Or-opt (ILS + Or-opt) for packing formation support systems to improve the efficiency of planning operations. This algorithm has the following two features. First, it uses Or-opt to create a neighborhood for a local search. Second, uses exchange neighborhoods when repeating the ILS to achieve diversity in the search.