- Qingmu Deng
- Sam Daitzman
- Oscar De La Garza
- Kayla Brand, Wellesley College
- To support materials scientists developing novel materials by providing advanced data science techniques to accelerate research.
- To develop user-friendly software tools that enable more scientists to use advanced techniques.
- To provide scalable training for scientists to adopt advanced data science techniques.
Materials development is a slow, often wasteful process. Scientists use their intuition to help guide development but, like Edison developing the lightbulb, there is a large amount of random guesswork. Accelerating materials development could reduce waste and lead to improved technologies, such as solar cells and batteries to drive electrification and combat climate change.
The work I do is on the human side: treating scientists as users and stakeholders, and developing technologies to help them develop novel materials faster. By working directly with material scientists, I help individual investigators accelerate their research. Students who work with me on this effort learn how to use machine learning to support materials development.
Through those partnerships, I gain generalizable insights into scientists' needs. I implement these insights in the software package Grama, creating a reusable software toolkit co-designed with scientist partners. Students who work with me on this software learn how to do rigorous open-source software development, and get to design and implement advanced algorithms for data science.
To scale these tools to a large audience I offer workshops on their use, most notably a recurring workshop sponsored by Georgia Tech's Institute for Materials. I am also writing a textbook to be published by Cambridge Scholar Press that will consolidate and teach best-practices for scientific modeling.
This research was made possible through a grant from Citrine Informatics, Toyota Research Institute and Wellesley College.