ML Disordered Materials Intern

At Toyota Research Institute (TRI), we’re on a mission to improve the quality of human life. We’re developing new tools and capabilities to amplify the human experience. To lead this transformative shift in mobility, we’ve built a world-class team in Robotics, Human-Centered AI, Human Interactive Driving, and Energy & Materials.

This is a Summer 2024 paid 12-week internship opportunity. Please note that this internship will be a hybrid in-office role.

The Energy and Materials Division at TRI is building tools to accelerate the design and discovery of new materials, fostering a transition to more sustainable mobility. Our research applies AI, data-driven methods, and automation to materials science, and spans the atomic to the device scales. Our projects often involve collaboration with scientists from universities and national labs. Interns will be involved in industrial research on topics of broader interest to the general materials science community, and several previous intern projects have resulted in peer-reviewed publications.

Machine learning models have improved dramatically in recent years as surrogate models for DFT, but lack predictive capabilities that can be applied to representations of materials with partial occupancies, i.e. disordered materials, which comprise well over two-thirds of the solid materials in the international crystal structure database (ICSD). We aim to create a machine learning model architecture that can be applied to disordered materials in the same way that m3gnet, matlantis, chgnet, etc.

have been applied to ordered materials. We then aim to benchmark the developed model towards the prediction of both functionality and synthesizability of disordered inorganic materials.

• Applicant must be pursuing a doctorate in materials science, chemical engineering, mechanical engineering, physics, applied mathematics, computer science, other engineering, or related field
• Experience in machine learning for solid inorganic materials, e.g. using or developing models like M3GNet, ALIGNN, CHGNet, etc
• Collaborative open-source software development, e.g. contributing to pymatgen, ASE or development of independent software projects hosted on e.g. Git Hub

• Experience using cluster expansions, special quasi-random structure (SQS), coherent potential approximation (CPA), or similar methods targeted at materials with partial occupancies
• Experience in DFT or other electronic structure methods
• Translation of theoretical predictions into laboratory-synthesized materials, ideally for applications in batteries or fuel cells

The pay range for this position at commencement of employment is expected to be between $45 and $65/hour for California-based roles; however, base pay offered may vary depending on multiple individualized factors, including market location, job-related knowledge, skills, and experience. Note that TRI offers a generous benefits package including vacation and sick time. Details of participation in these benefit plans will be provided if an employee receives an offer of employment.