I develop computer vision and AI models for spatial (image) data

I am a Data Scientist at Bayer's Crop Science R&D Division where I contribute to Digital Twin by developing deep learning models that leverage real time spatio-temporal data from satellite imagery to nowcast future traits. I have expertise in developing self-supervised and AI foundation models, energy modeling, nowcasting, remote sensing, and machine learning. 

Work Experience:

Data Scientist - Bayer R&D, June 2023 - Present

Research Scientist - Gro Intelligence, April 2023 - May 2023

Machine Learning Scientist (Post-Doc) - CIRA CSU and NSF AI2ES, January 2022 - March 2023

Research Intern - GeoAI group @Oak Ridge National Labs, Tennessee, June-August 2021

Research Assistant (Pre-Doc) - School of Computer Science & Engineering - NTU, Singapore, 2014-2015 

Education:

Ph.D. in Computer Engineering, University of Massachusetts AmherstSep. 2015 - Oct. 2021

M.S. by Research in Information Technology, IIIT- Bangalore (India), 2011-2014

B.Tech in Computer Science, MDU Rohtak (India), 2007-2011 

CV

Personal

I was born and grew up in India. I am married to Trapit Bansal and live in San Francisco, California with our daughters Yuki and Ayomi. 

Apart from research, I enjoy spending time with our daughters, cooking, doodling, playing board games and video games.

News 

December'23: Will be attending AGU'23 in-person in San Francisco, California. Please consider submitting to the session "IN028 - GeoAI and the need for Everything Everywhere All at Once" and reach out if you are planning to attend!

October'23: Will be attending "Google Geo for Good Summit" in-person in Mountain View, California. Feel free to reach out if you are planning to attend!

June'23: Thrilled to join Bayer R&D as Data Scientist working at the intersection of ML, satellite data and agriculture!

Super excited to join Gro Intelligence as Research Scientist!

Presenting two talks from the recent work at the 103rd AMS Annual Meeting in January'23

Released our latest work "Tools for Extracting Spatio-Temporal Patterns in Meteorological Image Sequences: From Feature Engineering to Attention-Based Neural Networks" on arXiv

Speaking at the Satellite Data for Energy Analysis and Policy at University of Wisconsin - Madison in October'22

Speaking at the Transformers for Environmental Science (ESST) Workshop in September'22

Join us at the NOAA 4th Workshop on Leveraging AI in Environmental Sciences in September'22

Join us at the Trustworthy Artificial Intelligence for Environmental Science (TAI4ES) Summer School organized by NSF AI2ES & NCAR from 27-30 June'22 . Registration is mandatory and free for students!

Our paper "A Moment in the Sun: Solar Nowcasting from Multispectral Satellite Data using Self-Supervised Learning" got accepted in ACM e-Energy'22 conference. Catch the virtual talk on 29th June'22 at e-Energy'22

Very excited to share that I have joined CIRA Colorado and NSF AI2ES institute as Machine Learning Scientist (Postdoc). 

Just released the final chapter of my Ph.D. thesis in our new paper

I successfully passed my Ph.D. thesis defense on 18th October'21 !! Yayyyyyy

Our paper "Self-Supervised Learning on Multispectral Satellite Data for Near-Term Solar Forecasting" got accepted in ICML Workshop on Tackling Climate Change with Machine Learning'21.

Super excited to be spending the Summer'21 researching at the GeoAI group, Oak Ridge National Labs!

I won third prize in the ECE 3-Minute Thesis (3MT)'21 Competition at UMass Amherst.

Serving as TPC member for FATEsys'21.

I passed my Ph.D. Dissertation Proposal titled, "Data-Driven Control, Modeling, and Forecasting for Residential Solar Power", at UMass Amherst.

I won second prize in the Lightning Talks Competition during the 2020 Energy Data Analytics Symposium! Check the talk here, and the media coverage on Twitter, LinkedIn, Facebook.

Check out the media coverage by UMass and UMass IPO on the Duke Data Analytics talk.

Our Paper "See the Light: Modeling Solar Performance using Multispectral Satellite Data", received honorable mention under best presentation award in ACM Buildsys'20 

Our Paper "See the Light: Modeling Solar Performance using Multispectral Satellite Data", Akansha Singh Bansal and David Irwin got accepted in ACM Buildsys'20.

Our Paper "Exploiting Satellite Data for Solar Performance Modeling", Akansha Singh Bansal and David Irwin got accepted in IEEE Smartgridcomm'20.

I am teaching an undergraduate first-year seminar course at UMass Amherst this Fall'2020 titled - "Staring at the Sun: An Introduction to Solar Energy". 

I am attending Grace Hopper Celebration (GHC'19) at Orlando, Florida - (October 1 - October 4' 2019).

I will be presenting our paper (On the Feasibility, Cost, and Carbon Emissions of Grid Defection) at IEEE Smartgridcomm'19 in Beijing, China - (October 20 - October 24' 2019).