Solution methods for the integrated production planning and scheduling in molded pulp industry

Abstract

This research project aims to propose specific solution methods for the integrated production planning and scheduling in molded pulp industry. In this industrial context, the production process involves stages as blending, molding, drying and packaging. The main production planning decisions consist of defining (i) which combination of molds should be attached to the molding machine (i.e. defining which molding patterns should be used), (ii) how long these molding patterns should be used and (iii) how they should be sequenced. The number of possible molding patterns that can be set up to the production line may be large to be defined in advance, so the production planning decisions are usually limited to just a subset of them. With the purpose of overcoming this limitation, which might lead to suboptimal production plans, this research proposes approaches for integrating decisions about generating molding patterns, production planning and scheduling. Besides the challenges involved in the integration of these decisions, another concern is to generate feasible molding patterns which, comply the operational and technical constraints of the production system.We proposed two mathematical formulations for addressing the integrated problem in molded pulp industry. The first formulation defines molding patterns based on settings defined in advance for the molding machine and integrates this issue to the production planning and scheduling. The second formulation includes the operational and technological features of the production system, generating feasible molding patterns and production plans simultaneously. These mathematical models were solved by CPLEX 12.5.0.0 for a set of instances based on data provided by a Brazilian molded pulp packaging company. Preliminary results showed that for specific scenarios, the proposed models could lead to the optimal solution in some instances. However, the performance of the models is limited for other realistic group of instances. Thus, this research projects aims to study solution strategies based on mathematical programming, reformulations and heuristics proceedings in order to solve effectively the integrated problem. (AU)