Tools Of Monte Carlo Simulation In Inventory Management Problems

Abstract

This paper describes the simulation model of supply chain and its implementation using general purpose tool and the simulation package. The output of Monte Carlo experiments taken both from spreadsheet formulas in Microsoft Excel and from graphical environment of Extend software was confronted and revised and then model was used to find the minimal inventory cost. A BRIEF DESCRIPTION OF THE INVENTORY MANAGEMENT'S BASIC CONCEPTS According to (Heizer and Render 2001) there are four types of firm’s inventory: • Raw material inventory, which are at command of the firm. The main purpose of maintain of these items is to eliminate supplier variability in quality, quantity or delivery time. • Work-in-process inventory, which are processed inside the firm. It means that changes has been made but not ended yet. • Inventories which are devoted to maintenance, repair and operating. It refers to assure current running of plants and devices etc. • Finished goods inventory: the completed products waiting for shipping. They arise because values of future demand are unknown. There are two types of models of inventory control in relation to demand for other goods: dependent (for instance portable FM receivers and batteries) and independent (for instance TV receivers and FM receivers). In our paper, we will study the issues related only to independent demand. The quality of implementation of inventory control policy is evaluated according to total inventory cost over certain time. Total inventory cost is composed of holding costs (also including insurance cost, handling etc.), ordering cost (also transportation cost, packing and setup cost, administration cost) and stockout cost (storage cost). The main purpose of model analysis is generally to minimize the total costs. One can assume sometimes that the demand is known or constant but most often this assumption have to be rejected and it is necessary to specify the probability distribution of demand. This leads to a probabilistic model. That sort of models are sometimes evaluated by service level, which usually can be measured at a number of points in the supply chain e.g. X service levels means that the products is available X percent of the time or X out of a hundred customers will buy the product. To reduce stockouts one can increase inventory in comparison with results from non probabilistic model. Size of safety stock depend on stockout and maintain costs, including ordering and storage cost. There are two patterns (approaches) of inventory controlling and scheduling – push (for instance MRP, based on production schedules, developed for production stages according to demand forecasts) and pull system (for instance just in time and kanban). We will present the simulation model of a pull system. It is based upon the direct and immediate ordering products or components by the customer from the supplier who delivers it at required time and amount. IMPLEMENTATION OF THE INVENTORY MANAGEMENT BUSINESS GAME Authors investigated the process of implementation of business game, based upon kanban approach which applies the probabilistic model of goods demand and time of fulfilling the order by the supplier. Our considerations were based on inventory management model from (Heizer and Render 2001) but very similar approach is presented in (Lawrence and Pasternack 2001; Jensen and Bard 2003). Guidelines of the Inventory Management Business