Applying Available-to-Promise (ATP) Concept in Multi-Model Assembly Line Planning Problems in a Make-to-Order (MTO) Environment

Abstract

We consider a multi-model assembly line production planning problem. We assume an environment where orders with several different model types with varying quantities are received by contract manufacturers in a Make-to-Order (MTO) environment. The models are similar enough in such a way that they share some common critical raw materials/parts and are to be produced on an already balanced multi model assembly line. Due to MTO and contracts there are significant costs associated with earliness and tardiness in addition to inventory and production costs capacity and other operational constraints. The challenge is to be able to make quick and accurate decisions regarding whether or not to accept an order and provide a due date along with raw material procurement and production plan that can be followed. This problem is closely related to Available to Promise (ATP) systems. Increased revenue and profitability are expected with the better management of ATP systems by reducing the amount of missed market opportunities and improving operational efficiency. This study aims to develop an effective solution method for this problem which will minimize earliness tardiness lost sales inventory holding FGI subcontracting overtime and raw material costs. We present the results of a detailed review of related literature. This study fills the gap in the literature of assembly line planning problems that covers Available-to-Promise by considering shipping decisions on critical raw materials required for production in a make-to-order environment. After determining a gap in the literature a novel mathematical model has been developed to solve the problem on hand. While this developed mathematical model offers an acceptable calculation time for problems where the production time required to meet the total demand does not exceed the total inhouse production time it does not offer a fast solution method for the problems where the total inhouse production time is insufficient to meet the total demand. For this reason a heuristic algorithm has been developed that provides faster results with near-optimal solutions for this type of problem. We present the results of our experimentation with both models.