Formulation Blending Recognition use Artificial Intelligence (AI)

Abstract

This project involves with artificial intelligence to predict the blending failure for the contact lens.  The formulation for the contact lens is Lotrafilcon formulation that consists of betacon macromer, TRIS, DMA, ethanol and irgacure. Each raw material in the formulation has a range of weight controlled by the Chemical Production that manually weigh by the operator. Then, the material is mixed together for about 6 hours but after 2.5 hours a sample is taken from the mixing material to be analyze by the Analytic Lab. But when the formulation undergoes FTNIR analysis, the mixture fails due to betacon macromere weight that does not equivalent with the manually weighted one. This project will use artificial intelligence to train a system with the existing data of the analysis from Analytic Lab and Chemical Production and predict a weigh data that will match the formulation specification and pass the FTNIR analysis.