Organizers:
- Hanxun Jin, Washington University in St. Louis
- Grace Gu, University of California Berkeley
- Sinan Keten, Northwestern University
- Guy Genin, Washington University in St. Louis
- Markus Buehler, Massachusetts Institute of Technology
Description:
The convergence of mechanobiology, biomaterial design, and ML/AI is catalyzing a paradigm shift in our ability to decode biological systems and engineer next-generation materials. The emergence of generative and physics-aware AI models, multi-agent AI frameworks, and self-improving computational platforms is transforming how we predict, design, and optimize biomaterials. These approaches enable real-time knowledge synthesis, autonomous discovery, and closed-loop integration of experimental validation.
This symposium will explore how cutting-edge ML and AI techniques—including multi-modal and graph-based reasoning, physics-informed neural networks, and self-reflective agentic AI—are revolutionizing biomechanics and biomaterials. We will highlight how AI-driven models not only analyze existing data but generate hypotheses, design simulations, execute experiments, and refine results in an iterative fashion.
We aim to convene leading researchers at the interface of ML, biomechanics, and biomaterials to address fundamental questions: How can AI-driven models unravel structure-function relationships in biological materials, including living systems, bio-inspired materials, and other engineered materials? What novel capabilities do multi-agent AI systems bring to biomaterials research? How can we best integrate physics-based simulations with experimental validation, for instance in a self-correcting loop that incorporates experiment, first-principles theory, and generative methods? We especially welcome contributions that showcase physics-aware AI, hierarchical modeling, and agentic learning systems that push the frontier of theory, experimentation and simulation.
Topics of interest:
- Multi-Agent AI and Automated Discovery: The role of agentic AI systems in designing and optimizing living and bio-inspired biomaterials through self-learning or autonomous experimentation.
- Physics-Informed ML for Complex Biological Systems: Integration of physical constraints and first-principles modeling to ensure mechanistic accuracy and interpretability.
- ML Applications in Biomechanics: Insights into tissue mechanics, cellular behavior, and organ-level modeling using AI-driven frameworks, including vision models for biomaterials, biology and other areas in engineering science.
- Materiomic AI for Biomaterial Design: Leveraging large-scale models for de novo biomaterial synthesis, self-assembling structures, and adaptive bio-inspired materials.
- Graph-Based AI for Structure-Function Mapping: Use of transformer-based models and graph reasoning to identify and optimize biological and biomimetic materials.
- Integrating AI with Experimental Validation: Strategies for bridging computational predictions with real-world testing, uncertainty quantification, and closed-loop material discovery.