Task-driven Automation Empowered by Multidisciplinary Engineering Tools – Principles and Case Studies from Solid-state and API Engineering Development Lab

As the complexity of the new chemical entities continues to increase, the urgency to expedite drug development amid constrained resources is becoming more critical. Task-driven automation, supported by multidisciplinary engineering tools, is revolutionizing workflows in solid-state and API engineering laboratories. Our lab focuses on understanding polymorphic behavior and designing scalable processes using thermodynamic and kinetic data. To shorted development timelines, we automate both physically demanding laboratory tasks and computationally intensive in-silico models. This requires not only expertise in solid-state chemistry and chemical engineering, but also proficiency in web programming, machine learning, and mechanical engineering.
We have effectively automated two lab tasks—XRPD characterization and solubility measurement—by replacing the most labor-intensive preparation steps with custom-engineered tools. These automation solutions have been implemented and validated across multiple projects. To further enhance data analysis and visualization, we built user interfaces using web technologies.
In addition, we developed a predictive tool leveraging machine learning to evaluate polymorph complexity, providing an indicator of a compound’s tendency to form hydrates or solvates based on its 2D structure and commonly used solvents.