User-Modified Artificial Intelligence Causal Prediction Model for Process Improvement Using Design Document Interface

Machine learning a artificial intelligence (AI) derives correlations between manufacturing variables from the data. These correlations can be classified into causal and noncausal relationships. In the design for manufacturing, only causal relationships are applicable because control based on causality is necessary to produce the desired product. Domain knowledge is required to confirm the causality and develop a predictive model based on causal relationships. However, domain experts often lack the AI-related knowledge necessary to develop such models, including the skills in AI, coding, and data mining. To overcome this challenge, we developed an AI system that leverages a design document interface (DDI), allowing domain experts to easily create AI models tailored to their tasks even without extensive AI expertise.