Improvement of inventory policy for furniture fittings manufacturer

Abstract

The objective of this research is to improve the management of inventory in a furniture fittings manufacturer so that shortage and overstock could be reduced. Based on our initial investigations, such inventory problems were caused not only by inefficient production planning but also a lack of inventory control system that linked those plans with the stocks of raw material required. To better address the issues, two different models have been therefore proposed, namely more accurate demand forecasting models for products with different demand patterns and inventory control models for raw materials of those products.

Regarding the development of more accurate demand forecasting models, we first classify the products into six groups based on value of items and lead time. We then select ten products in each of three subgroups for further analysis, where a number of demand forecasting models have been applied to each product group according to their demand patterns. The study reveals that the Autoregressive Integrated Moving Average (ARIMA) model provides the least forecasting errors, as measured by Mean Squared Error (MSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE), for products in Groups A and B, while Croston’s model is the most appropriate forecasting model for products in Group C as it provides the least forecasting errors as measured by Mean Absolute Scaled Error (MASE).

We also evaluate inventory policies and determine stock levels of the company’s raw material under two different systems, i.e. continuous and periodic reviews. We find that, with proper inventory control model, the company could save up to 32% for special raw materials/components under continuous review and about 43% for common raw materials/components under 30 days periodic review.