Sustainability evaluation of alternative part configurations in product design: weighted decision matrix and artificial neural network approach

摘要

Recently, sustainable products have been demanded by world legislation not only to generate profits, meet consumers' needs, or reduce adverse impacts on the environment, but also considering all aspects of economic, societal, and environmental. Numerous sustainable product designs have been introduced by considering sustainability in developing product concept; however, the sustainability issues considered during the evaluation of different product configurations are rather limited. In response, this study proposes a systematic approach to evaluate sustainability of configuration design alternatives based on weighted decision matrix and artificial neural network. Weighted decision matrix is used as a platform to deal with inconsistency data based on a point scale and generate ratings based on the score given. In this phase, configuration designs are measured at an interval scale and the score is derived using a nine-point Likert scale, which is used to determine the score of the configuration design with regard to the sustainability criteria compared with other alternatives. Then, artificial neural network is utilized to estimate the sustainability performance in a single value, named Weighted Sustainability Score by aggregating the generated ratings using a trained network. A case study of an armed chair is conducted to illustrate the proposed approach in detail. The accuracy of the proposed approach on the result for environmental evaluation is validated by environmental-based commercial software. The results demonstrate that the proposed approach provides similar decision with the commercial software in ranking the alternative part configurations with regard to environmental consideration. Consequently, the study shows the effectiveness of the proposed approach in a part of evaluating environmental aspect, and would do the same for systematically evaluating sustainability elements.