MATHEMATICAL MODELLING OF PRODUCTS ALLOCATION TO CUSTOMERS FOR SEMICONDUCTOR SUPPLY CHAIN

摘要

Where demand outstrips supply, there will result in shortages to end customers. In such a case decisions need to be made of how to allocate supply to customers. Customer satisfaction requires accurate order promising that leads to better cooperation, as well as trustable orders and forecasts from customers. As a result, customer satisfaction through a trustable promising system leads to more accurate planning for production. In this regard, modern Advanced Planning Systems (APS) provides allocation planning to customers’ orders based on “Available To Promise” (ATP). Lack of supply, escalation, and excess demand are propelled by competitive plant capacity, dynamic behaviours of ATP, orders, and demand forecasts in demanding industries like semiconductor manufacturing. When demand exceeds supply, APS needs the support of experts (human intervention) about the time and amount to be allocated to customers. This feature of APS keeps the flexibility of planning to find feasible optimal decisions regarding allocations. In this paper, we propose a mathematical model for the optimization of ATP allocation to customers, where demand exceeds supply, which will be presented as a decision support tool to analyse allocation scenarios. The objective of the proposed mathematical model is maximizing customer service level which is directly related to customer satisfaction while keeping a maximum of stock. The model is being developed from a case study of a European semiconductor supply chain with a sales office in Ireland. In this case study, support will be provided to allocation managers that allows flexibility solutions to be developed. The obtained results have validated the proposed multi-objective mathematical model.