The following is an excerpt from “6 Hopkins innovations featured at the World Changing Ideas Summit,” published by the Hopkins Bloomberg Center. Click here for full coverage of the World Changing Ideas Summit.
The materials that go into protecting a person, a structure, or even the planet must be able to withstand the conditions they will experience—immense speeds, extreme heat, and very high pressures, all applied in the blink of an eye. Engineering design for protection has traditionally been done by simply choosing from among already available materials, but very few existing materials can survive these conditions. Creating entirely new materials to specifically meet that function has been difficult, because the range of possible materials is very large, and testing under extreme conditions is slow and very expensive.
But K.T. Ramesh, Alonzo G. Decker Jr. Professor of Science and Engineering, is using AI and automation to help quickly and cheaply design new materials for these kinds of applications. His lab has developed an approach that speeds up the testing of materials by a factor of 1,000, while lowering the cost of each test a hundredfold. AI is then used to learn from this huge amount of data to design new materials, such as the skin of a spacecraft that is entering the atmosphere at high speeds. This is helping make new material development both possible and more affordable.
“So now you can change the question from ‘What do I have?’ to ‘What do I want?’” he said.
Ramesh is a part of the AI for Materials Design Laboratory at Johns Hopkins, which is designing methods and models to discover new materials quickly and cheaply with a focus on situations like soldier protection and hypersonics.
Ramesh’s work also leverages machine learning to more easily predict how a material or structure will respond under extreme conditions. Simulations that were once slow and expensive because of the computational power they required now take a fraction of the time with AI. This, he added, has the potential to transform how and what is built in the future.
“I think we’re transitioning from a society that is mostly thinking about how to understand things as they are, into one that is able to make the things we want and make them at a relatively low human cost,” he said.
