Rishi E. Kumar

Postdoctoral Researcher

About Me

Hi, my name’s Rishi and I am developing software for high-throughput material discovery and optimization at Lawrence Berkeley National Lab. I bridge hardware, software, and materials expertise to automate material science research. Additionally, I have developed spatially resolved metrology techniques for benchtop and synchrotron characterization of photovoltaics.

I am most skilled in: Materials Characterization, Lab Automation, and Python.

Get my CV here.

Code for Lab Automation

PASCAL

The Perovskite Automated Solar Cell Assembly Line (PASCAL) at the Fenning Lab in UC San Diego. Spin-coats and characterizes optoelectronic thin films, constructs solar cells, and executes closed-loop active learning optimization campaigns. A constant work in progress!

Check it out in action here and here!

Roboflo

Task scheduler for any system with coordinated workers. The original use case is for the Perovskite Automated Solar Cell Assembly Line (PASCAL) in the Fenning Lab at UC San Diego, where a robotic arm moves small glass slides between stations to perform experiments.

MixSol

Pipetting planner for efficient combinatorial mixing of solutions. Often we want to interpolate a few stock solutions into many target mixtures. If some of these mixtures require only a tiny amount of a stock solution, the minimum volume for our pipette may limit our ability to make this solution without a serial dilution. Mixsol searches for mixing sequences that use only other target solutions as stepping stones to reach these difficult mixtures, minimizing waste.

Experience

Hacking Materials at Lawrence Berkeley National Lab

Postdoctoral Researcher

2022 - Present

hackingmaterials.lbl.gov

Developing software for automated experimentation platforms. This ranges from backend, hardware orchestration to AI-driven decision making in closed-loop experiments.

Fenning Lab at UC San Diego

PhD Candidate

2017 - 2022

fenningresearchgroup.com

Worked on a variety of research topics related to photovoltaics (PV), with an emphasis on doing faster science. Built a robot to do high-throughput compositional search and/or process optimization for halide perovskite thin films. Used machine learning to guide the robot to do the right experiments. Developed a method to measure water in silicon PV modules without taking them apart, used that to study local effects of water on performance loss. Did a lot of synchrotron characterization of halide perovskite composition and structure.

Nanocomposix

Research Associate II

2015-2017

nanocomposix.com

Made pyrophoric confetti to mimic the heat signature of specific aircraft to distract advanced heat-seaking arms. Controlled the size and shape of silver nanoparticle to selectively absorb common laser wavelengths for skincare applications. Led the scaleup of this nanoparticle synthesis from 5 mL to 55 gallon batch sizes.

Kyocera North America

R & D Engineer

2014-2015

americas.kyocera.com

Determined location and cause of manufacturing defects in ceramic packaging for integrated circuits. Optimized process conditions for new soldering pastes used for die-attach.

Education

University of California San Diego

PhD Materials Science and Engineering

2017-Present (estimated spring 2022)

University of California San Diego

BS Materials Science and Engineering

2010-2014