Toward Automated Qualitative Supply Chain Diagnoses in Engineering-to-Order Environment
Abstract
Engineer-To-Order (ETO) environments are more and more numerous in Supply Chains. Most of the concerned companies use continuous improvement methods such as Lean Engineering, Agile Management or Thinking Processes from the Theory of Constraints to stay competitive. All these methods start with a diagnosis step and their relevance is highly depending of the quality of this step. However, literature analysis shows that they are often time-consuming and inconsistent. One of the key issues of these weaknesses is that they do not consider enough the qualitative dimension of the ETO context. Consequently, we suggest in this paper a decision support system able to manage a rapid and relevant qualitative ETO diagnosis based on the Thinking Processes paradigm. This system is designed around a generic reality tree knowledge base and a set of dedicated inference rules. A real industrial application case is described and discussed. In this case study of an ETO oil&gas company, the diagnosis process has been reduced from 15 days to 1 day using this tool with the same starting conditions of the classical diagnosis.