Implementing a Digital Transformation in Process Chemistry: Integrated Automation, Machine Learning, and Real-Time Analytics from Lab to Pilot Scale

Abstract

Implementing a digital transformation in the Process Research and Development (PR&D) phase of an active ingredient offers significant opportunities to accelerate the journey from laboratory to manufacturing scale. Here, we report on three distinct initiatives undertaken at Syngenta to address concrete bottlenecks at different PR&D stages. First, Bayesian optimisation enabled efficient navigation of large reaction spaces with minimal experimental effort. Second, laboratory automation combined with multilinear calibration reduced hands-on time by ~85% and laid the foundation for autonomous closed-loop optimisation. Third, advances in Process Analytical Technology (PAT), including improved Multivariate Curve Resolution algorithms and modular Python-based pipelines, enabled real-time reaction monitoring in challenging industrial settings.