An intelligent benchmark-based design for environment system for derivative electronic product development

In recent years, the destruction of the ecological environment and the exhaustion of natural resources have become increasingly severe. The demand for environmental protection has attracted worldwide attention. If a company cannot address changes in ecological law, its products will no longer have access to world markets. For this reason, companies should introduce the concept of design for environment (DfE) to the product development process to alleviate the impacts of a product on the ecological environment throughout the product life cycle. To help companies effectively meet the challenges cited above, this study develops an intelligent benchmark-based DfE system for derivative electronic products. The architecture of the intelligent system includes three parts. In the first part, a back-propagation neural network (BPNN) is used to create the analysis of approximate life cycle inventory (LCI). The first part is utilised to estimate the amounts of hazardous chemical substances produced by the material extraction, manufacturing and usage phases as well as the amounts of energy consumed by the production and usage phases. In the second part, data envelopment analysis (DEA) is applied to develop the models of benchmark-based DfE evaluation for a product and a component or part. The models are used to identify which hazardous chemical substances are excessive and to calculate the needed reductions in the quantities of these hazardous chemical substances. The third part of the intelligent DfE system is used to generate analysis and suggestions for improvement in the design of parts or components. The analysis and suggestion model utilises the second-part analytical results and a BPNN to identify concrete directions for improvement and to generate suggestions that can be used to achieve the ultimate goal of DfE. (C) 2012 Elsevier B.V. All rights reserved.