| Scope |
AI/ML is rapidly transforming the paradigm of materials research and manufacturing, particularly through self-driving laboratories that integrate real-time experimentation and data-driven decision-making to accelerate materials optimization and scientific discovery. This symposium will focus on recent advances in AI-driven autonomous experimentation and closed-loop, expert-amplifying materials discovery, from research laboratories to industrial applications. Topics will include multimodal correlative analysis across disparate experimental and computational modalities; autonomous theory-experiment coupling; active-learning frameworks, offline reinforcement learning, and diffusion-guided experimental planning. Particular emphasis will be placed on emerging developments such as on-the-fly knowledge injection, where live experiments can adapt continuously to external user feedback, evolving scientific hypotheses, and uncertainty estimates. In addition, the symposium will explore the transition from isolated self-driving laboratories toward interconnected, cross-facility autonomous research platforms that link user facilities, university labs, and industrial testbeds. The symposium aims to bring together researchers from materials science, AI/ML, and high-performance computing to discuss implementations that will advance autonomous scientific discovery and manufacturing while broadening access beyond expert-only operation.
Key Topics Included (but are not limited to):
• AI/ML for multimodal correlative analysis and autonomous theory-experiment coupling
• Closed-loop active learning, offline reinforcement learning, and diffusion-guided experimental planning
• On-the-fly knowledge injection, human-in-the-loop guidance, and expert-amplifying autonomy
• Autonomous and AI-guided manufacturing
• Materials discovery through generative AI and agentic AI systems for safe, uncertainty-aware experimental decision-making
• Cross-facility & multi-site autonomous platforms, including lessons learned from historical experiments, failed trials, and facility experimental trajectories
• Data infrastructure for autonomous science, including experiment metadata, FAIR data practices, and workflow interoperability |