Managing Uncertain Inventory in Supply Chain with Neural Network and Radio Frequency Identification (RFID)

Managing Uncertain Inventory in Supply Chain with Neural Network and Radio Frequency Identification (RFID)

CKM Lee (Nanyang Technological University, Singapore), Ng Wenwei Benjamin (Nanyang Technological University, Singapore) and Shaligram Pokharel (Qatar University, Qatar)
DOI: 10.4018/978-1-60960-585-8.ch010
OnDemand PDF Download:
No Current Special Offers


Demand uncertainty leads to fluctuations in inventory position at each echelon of a supply chain causing bullwhip effect, which can lead to significant cost and loss of efficiency and waste of resources. One of the aspects that can reduce potential bullwhip effect is the sharing of real time information for which the recently mass produced Radio Frequency Identification (RFID) can be of great value. The use of RFID technology can also help in increasing the visibility of the flow of goods and material, keeping track of the location and quantity at each distribution centre and warehouses. This will also help in the periodic and near real time optimization of inventory level of goods and material. The data collected with RFID can be analysed in artificial Neural Network (NN) to forecast the future demand. In this chapter, a framework is proposed by combining RFID with artificial neural network so that lean logistics can be realized in the supply chain.
Chapter Preview

The Trend In Rfid Based Modelling

The quantitative forecasting method can be generally classified as time-series model and causal model. Time series model includes naïve forecast, moving average (MA) and autoregressive (AR), exponential smoothing (ES), double exponential smoothing (DES) and triple exponential smoothing (TES). In causal forecasting method, a forecast of the quantity of interest (the dependent variable) is obtained by relating it to one or more of independent variables such as lead time, customer satisfaction and product reliability.

Saygin (2006) has proposed the use of forecast-integrated inventory model under simulation environment and showed that the model combined with RFID can produce the desired level of system performance in terms of lowering manufacturing costs and inventory levels. RFID can also be used for managing ordnance inventory with multiple attribute utility structure to realize the non-cost benefit such as safety, manpower issues and ROI analysis (Doeer et al., 2006). Choy et al. (2009) use a case based reasoning method for selecting material handling equipment and the shortest order picking path by using RFID. The use of RFID for inventory accuracy was examined by Uçkun (2006) and Uçkun et al., (2008) and showed that RFID investment can be highly beneficial for sharing inventory information within entities in a supply chain.

Complete Chapter List

Search this Book: