Early Warning System Design for WEEE Reverse Logistic Network

Abstract

The increase of environmental concern as a result of corporate citizenship spreads the applications for collecting end-of-life products to a broader extent. This trend raises the issue of reverse logistics (RL), one of the major challenges in sustainability. One of the greatest barriers for successful RL is the difficulty of developing an accurate system to forecast the amount of product returns. Advanced techniques such as learning systems are proven very helpful for increasing the performance of forecasting methods. This article proposes an “early warning system” for waste collection operations in the electrical and electronic equipment industry. The main goal is to develop a supportive system for manufacturers and authorized organizations that provides foresight about their potential to reach the target values proposed by environmental regulations. The proposed forecasting system is based on an artificial neural network (ANN) model with five basic factors affecting the amount of product return: sales amount, number of houses, electricity consumption, the GINI coefficient (coefficient showing income distribution inequality) and population density. An application of the system is shown for Marmara Region, Turkey, and the compliances of all the big cities in the Marmara Region are checked for target values. The researchers' findings show that only five of eleven cities will be successful at fulfilling the required target e-waste values addressed by WEEE regulations.