The production of pharmaceuticals generates a huge amount of data; Sanofi (Frankfurt-Höchst) uses machine learning to extract further insights for the purpose of optimization of all quality assurance processes. Using innovative approaches like those is part of Sanofi`s Digitalization Strategy. Sanofi is dedicated to supporting people through their health challenges.
Two student teams from Goethe University of Frankfurt conducted an innovative data analytics projects for Sanofi in cooperation with the chair of information systems (Prof. Dr. Minor) and Campana & Schott. The student teams were assigned to find insights in raw production data that incurred during production of plastic components for application devices that exceed simple specification deviations.
“The cooperation with universities helps us to incorporate innovative approaches in a scientifically sound way into our processes,” says Galina Hesse, Head of Digital, Engineering & Innovation, with focus on digitalization of production. “We actually benefit from the lack of experience in the pharma domain of the students, to approach this challenge without any bias.”
Both teams discovered great potential: using appropriate algorithms to predict the probability of rejects based on sub assembly data avoids subsequent assembly steps and therefore minimizes rejects. “Both cases revealed the need for a proper data strategy to get the most out of the data,” says Galina Hesse. The requirements to that data strategy are as important and complex as the results of the data analysis. During the project both teams compared various and machine learning methods and algorithms. The provided results will be used as a basis for a production-ready implementation. For Sanofi, this is a step on their way towards a higher degree of digitalization in production processes.
