(635c) Analysis of Industrial Sustainability Under Uncertainty Using Design of Experiments Technique and Fuzzy Logic Method

Complexity and uncertainty are the main features of industrial sustainability problems. These have made analysis and strategy development activities mostly scenario based, where the sustainability improvement factors and enhancement strategies are heuristically generated. There is a clear need for developing more systematic methodologies that are effective in talking sustainability problems of any size.

This work introduces a system analysis approach by integrating a Design of Experiments (DOE) technique and a fuzzy logic method that can be used to tackle industrial regional sustainability problem under uncertainty and to derive system solutions for sustainability enhancement. Due to its ability of assessing the mean effects between factor levels for a system response, DOE techniques are most suitable for generating a series of significance tests that are needed for checking the relations of the improving factors and their interactions with sustainability response of a given system. It is found that fuzzy logic may be used to extend conventional DOE techniques to deal with the uncertainties associated with data, information, and knowledge.

In this work, we employ fuzzy numbers and intervals handling techniques when applying the DOE. The basic algorithm of the proposed approach is structured in the following way. First, a fuzzy-logic-based multilayer sustainability assessment is designed for a given industrial system problem. With carefully defined fuzzy sets for the three aspects of social sustainability, the sustainability status of the zone can be appropriately determined under uncertainties of various types. Second, a number of potential sustainability improvement factors are introduced and a DOE technique is applied to analyze the sustainability problem by implementing a certain number of statistically designed trials. In each trial, a combination of different levels of the potential improvement factors is set as an input to the industrial system, and the sustainability status of the system is obtained by a fuzzy logic based multilayer sustainability assessment, which is an output response. The data of all trials are used to quantify the level of significance for each factor and their interaction to the sustainability status of the industrial system. Based on the quantitative results, we can readily separate significant factors and interactions from those insignificant ones, and gain a better understanding of the system. The approach can provide a scientific guidance in a holistic way for system sustainability enhancement. For instance, only those significant factors and interactions should be kept in consideration of sustainability enhancement, and limited resources should be distributed systematically on each factor based on the levels of significance and their interactions.

The main advantage of the introduced methodology is its capability of effectively and systematically analyzing industrial system problems with highly complexity and inherent uncertainty, and identifying and quantifying significant factors and their interactions that affect the sustainability enhancement. The methodological efficacy will be demonstrated by tackling a sustainability problem associated with automotive manufacturing.