(103b) Application of Functional Data Analysis (FDA) for Process Optimization

Functional Data Analysis (FDA) is an emerging field in statistics that allows for characterizing data over a continuum. Recently, FDA has gained a lot of attention in the literature and it has been successfully applied to data across multiple businesses in Dow R&D, but it hasn’t been applied to manufacturing data to this point. This motivates the interest of applying FDA in manufacturing applications, where analyzing time-series data is increasingly more common, especially in modeling and optimization of batch processes. Previous work had revealed some technical challenges that needed to be addressed to apply FDA to analyze manufacturing data. In this presentation, we will demonstrate how to deal with such gaps and apply FDA to manufacturing data for process optimization, deliver a comparison of FDA with existing methods, and recommend optimum conditions for applying FDA in modeling and optimization.