Publication Date
Advisor(s): Alexander Rothkopf
Topic(s) Covered:
  • Risk Management
  • Optimization
  • Manufacturing
  • Forecasting

The contract manufacturing industry is growing and shifting from standard products to highly customized engineer-to-order (ETO) products. Different from standard products, ETO orders have more production process uncertainties because their design specifications and production process can be changed after the orders have been accepted. Such uncertainties increase production costs, the risk of late delivery and associated penalties, exposing the contract manufacturers to profitability decrease. Since every ETO production is unique, companies cannot rely solely on historical data to ensure accurate planning. The goal of the project is to solve the aggregate production planning problem of ETO orders. Usually, uncertainty is mitigated by keeping inventories. Such safety stock, however, does not work for customized products as they are usually a one-time purchase and therefore, cannot be kept on a regular basis. For customized production, buffer time and capacity can be used against process uncertainty. We formulate an Aggregate Production Planning (APP) model as deterministic, multi-product, multi-stage, and multi- period linear programming (LP). It minimizes the total production cost by balancing the in-house production, inventory holding, outsourcing, overtime hours cost, and backlogged orders penalties. Cost drivers for total production cost are analyzed for multiple scenarios with different production times. We then calculated the buffer capacity and performed constraint sensitivity analysis using the shadow price method. Based on the data analysis, we make recommendations for the sponsor company for planning horizon that we model: add an employee for one production stage and remove an employee from another, use 7% buffer capacity for the base plan to minimize the total production cost for the set of all possible scenarios, use a combination of hiring, overtime hours, and outsourcing. These recommendations lead to 12.32% cost reduction compared to the cost if the company does not use an aggregate planning model and the recommendations. Moreover, we formulate an aggregate production planning approach for the sponsor company to use in the future to ensure an optimal plan with the minimum total production cost.