UROP Openings

Have a UROP opening you would like to submit?

Please fill out the form.

Submit your UROP opening

Developing a solar cell simulator based on JAX adsa


Term:

Fall

Department:

QI: MIT Quest for Intelligence

Faculty Supervisor:

Giuseppe Romano

Faculty email:

romanog@mit.edu

Apply by:

September 15th

Contact:

romanog@mit.edu

Project Description

Computational science is playing a prominent role in developing technologies for climate change mitigation. In fact, the predicting power of first-principles calculations, combined with the ever-increasing computing capability, enables fast material and device screening. To further boosting inverse design, novel methods are tapping into machine learning, where some concepts are readily usable. For example, once the a surrogate model based on artificial neural networks has been developed, the well-known backpropagation algorithm can aid the material search Within this context, our group is developing a solar cell simulator based on JAX, a novel Python-based framework which facilitates hardware-agnostic code and automatic differentiation. The aim of this tool is to provide an effective solver which can be seemlessly integrated in ML-based inverse design workflow. The UROP will help develop such a tool and use it for building a surrogate model based on artificial neural networks.

Pre-requisites

Fluency in Python and basic of calculus are required.