Subra Suresh

Abstract:

This work deals with our recent results from experimental, computational modeling, and data analytics of engineered and biological materials in three broad areas: materials science, plant science, and medical science. We show through examples and case studies how the appropriate combinations of experimental observations, two-dimensional and three-dimensional computational modeling, and images, as well as multi-fidelity data can be combined with physics-informed neural networks and biomimetics to improve materials design, and predictions of their properties and performance. A processing route to produce sustainable and nature-derived materials is presented whereby the building blocks can be tailor-made to produce digitally modulated structures, soft robotic components, and biocompatible substrate materials for wearable devices. For biomedical applications, novel approaches that integrate microfluidic platforms with static and dynamic data and images from clinical settings are also discussed to demonstrate how deep learning approaches can offer new possibilities to improve patient outcomes in disease diagnostics, therapeutics, and treatment. Specific cases considered here include: metallization of nanoscale diamond for tunable electronic properties; design of plant-based materials for soft robotics and sustainability; extraction of mechanical properties of materials through instrumented nanoindentation and multi-fidelity machine learning algorithms; and artificial intelligence velocimetry to probe diabetic retinopathy and blood disorders.