Challenge:

Process chemistry teams were eager to leverage machine learning to accelerate experimentation, but they faced significant barriers to adoption. Existing optimization tools lacked user-friendly interfaces, making them inaccessible to many scientists. Without a structured way to define and manage ML-driven experiment campaigns, teams relied on manual tracking of metadata, inputs, and results, slowing iteration and introducing inconsistencies. Additionally, experimental workflows were fragmented, with no integration between lab automation equipment, cloud-hosted ML models, and analytical instruments. The absence of a cure, real-time infrastructure for interacting with AWS-based services further limited the potential for scalable, intelligent experimentation.

Solution:

To overcome these challenges, the Advanced Software team developed a robust, SCADA-integrated dashboard platform tailored to the needs of process chemists. Built within Ignition SCADA, the dashboards provided intuitive, step-by-step interfaces for defining campaign metadata, experimental variables, and optimization objectives. A unified view allowed users to monitor model-suggested experiments, input results, and visualize outcomes – all in one place. The backend was powered by Microsoft SQL Server, ensuring reliable storage of campaign data and dashboard configurations. Secure API integrations enabled seamless communication with AWS-hosted Bayesian optimization models, developed in collaboration with the internal data science team. The platform also connected directly to lab automation hardware and LC/MS parsing workflows, creating a fully integrated experimentation ecosystem.

Results:

The deployment of the ML-guided experimentation dashboards transformed how process chemists conduct optimization campaigns. Scientists can now manage their workflows through a centralized, user-friendly interface that bridges the gap between machine learning, automation, and analytics. The platform has made experimentation more traceable, repeatable, and scalable, empowering teams to run more campaigns with greater confidence and less manual effort. By laying the groundwork for closed-loop, autonomous experimentation, the solution not only boosts experimental throughput but also enhances decision quality. This initiative marks a significant step toward smarter, faster, and more efficient R&D operations.

See how our team can assist?