This article focuses on the use of NO-CODE tools and artificial intelligence to extract these parameters and apply them to root-cause analysis in a product quality evaluation during the manufacturing of uranium pellets.
The main objective of this article is to demonstrate the commitment to continuous improvement through the potential of NO-CODE tools (Power Automate, AiBuilder, ChatGPT) in process automation, specifically for the automatic extraction of parameters from UO2 powder certificates. Prior to implementing this automated workflow, these data were inaccessible, which limited their potential. This new approach opens a wide range of possibilities in areas such as trend analysis, improvement teams, and enhanced process capabilities.
Through iterative use of ChatGPT, a Python script was developed to automate and periodically copy new files detected in a OneDrive (cloud) folder. Power Automate identifies these new PDF files and using AiBuilder, which is an artificial intelligence model pre-trained with various examples of certificates, extracts the relevant data and transfers them to another cloud repository.
This article illustrates, through a practical example, the value provided by this automatically generated data source. Its use enabled the root-cause analysis of a product quality event caused by specific characteristics of the UO2 powder used. Before the development of this workflow, such analysis would have been extremely challenging due to the inaccessibility of historical data.
