Inventory Management Efficiency Improvement Using Demand Forecasting Methods

Inventory management costs make up to 50 percent of logistics costs in the company. Therefore, the function has the highest potential for cost reduction. Inventory management optimization is defined through the quality of forecasting estimates pertaining to demand for products. Classic forecasting methods do not produce adequate forecasting estimates in situations when demand is intermittent or erratic. Lack of trend and limited number of observations of consumption turn the usage of analytical or probabilistic methods ineffective. Application of uncertainty theory developed by Liu, formalization of expert estimates processing in terms of demand forecasting and usage of uncertainty theory within EOQ model allows to obtain accurate demand forecasting estimates and calculate stock parameters. Testing of the suggested model is performed within the case of spare parts warehouse of the production facility. Inventory classification based on product volume attributes allows to determine the most critical positions in terms of storage volume reduction. The problem is relevant for the warehouses which are the property of the production company and have the limited amount of storage space. Algorithm for solving the problem of inventory parameters calculation for different nomenclature positions based on the number of observations for each of them is suggested. Provided the number of observations is sufficient, it is possible to present demand through Poisson distribution: in case the number of observations is small or there is none is it prudent to use uncertainty theory.