E-learning Intermediate

Automated Experiment in Materials Synthesis and Characterization

Provider The American Ceramic Society

Course description

This on demand cutting-edge course introduces materials scientists and ceramics researchers to machine learning-driven experimental automation, requiring no prior ML/AI experience. The course bridges the gap between traditional experimental methods and modern automated approaches.

Students will explore the transformative potential of automated experimental workflows, from single-tool automation to cloud-based laboratory systems. The curriculum covers essential techniques including Gaussian processes, Bayesian optimisation, hypothesis learning, and deep kernel learning for structure-property discovery. Advanced topics include manifold learning, variational autoencoders, and multifidelity approaches for orchestrating multiple characterization tools.

The course emphasises practical applications in materials synthesis and characterization while addressing real-world considerations such as cost-benefit analysis and human-in-the-loop experimental design. Participants will gain insights into explainable automated experiments and learn how to design reward-driven workflows that optimize experimental efficiency and discovery potential.

In this course, you will learn about:

Current opportunities for automated experiment in synthesis and characterization: from single tools to cloud laboratories
Objectives and rewards for automated experiment
Reward driven workflow design, orchestration, and execution
Gaussian processes and Bayesian Optimization
GP and BO in real world: cost and value
Bayesian Inference and structured Gaussian Processes
Hypothesis learning
Manifold learning, variational autoencoders, and encoders-decoders,
Deep Kernel Learning and structure-property discovery
DKL, explainable automated experiments, and human in the loop AE
Multifidelity structured GP for co-orchestration of multiple tools
Future perspectives

Machine Learning-Driven Experimental Design - Emphasizes practical implementation of AI/ML tools for experiment automation
Advanced Data Analysis and Pattern Recognition - Focuses on sophisticated analytical techniques for materials discovery
Automated Laboratory Integration and Orchestration - Highlights system-level skills for coordinating complex automated workflows

This course is designed for those who do not have prior machine learning and artificial intelligence experience, and may be materials scientists, engineers and researchers in ceramics and experimental materials who want to harness machine learning for automating synthesis and characterisation. It is also valuable for graduate students, lab technicians and R&D professionals aiming to accelerate materials discovery by integrating advanced automated tools and data-driven workflows.