Sasmita Sahoo, PhD

Senior Research Scientist, BrdgAI

(Digital Agriculture, Machine Learning, Geospatial Modeling, Hydrology)

I am a Research Scientist working on developing AI and machine learning solutions for digital agriculture and geospatial modeling. My work revolves around the agriculture-climate-water nexus, with primary focus on developing computational and geospatial models for addressing challenges related to food security, water resources and climate change.

More specifically, my research expertise includes:

  1. Building and deploying machine learning and process-based models to quantify the impact of environmental- and human-induced stresses on agricultural and water resources system.

  2. Developing and implementing custom models and tools based on sensors, satellite imagery, and ground observations to meet food security goals.

  3. Processing and analyzing large-scale, high-resolution geospatial data sets to draw insightful conclusions from them for future projections and decision making.

Previously, I was a Research Scientist at Michigan State University & Great Lakes Bioenergy Research Center. I was involved in developing and implementing data science and machine learning models in variety of applications, including:

    1. Integrating satellite and drone imagery, machine learning algorithms and process-based crop models to understand agricultural systems, to predict the impact of climate, soil and management on crop yield, nutrient uptake, and water use efficiency, and ultimately, to inform management decisions for sustainable agricultural production. Overall goal was to enable smarter solutions for precision agriculture in a resource-constrained and rapidly changing world.

    2. Developing and applying image classification and multi-level (supervised and unsupervised) machine learning algorithms for cloud and shadow detection in satellite imagery.

    3. Timeseries prediction of vegetation health from satellite imagery using historical change, geospatial data, soil and topography information.

    4. Integration of process-based crop models and machine learning models for large-scale yield and soil organic carbon prediction using soil, historical weather, and farm management data.

    5. Developing algorithms for canopy cover estimation and plant growth prediction from greenhouse (leafy greens) images using image segmentation and machine learning.

    6. Mapping crop growth, spacing and vegetation health from UAV imagery using image segmentation and edge detection algorithms.


Prior to joining MSU, I was a data scientist (Machine Learning) at Croptix, State College, a company that builds smartphone-based sensors for early disease detection in agricultural crops. I was involved in: 1) processing sensor data and developing machine learning models to detect early drought stress, nutrient (nitrogen, phosphorus) deficiency, and disease pressure in agricultural crops. ML models implemented for classifying spectral features are weighted k-nearest neighbor, decision tree, linear and quadratic discriminant analysis, support vector machine, ensemble models, and artificial neural network models. 2) developing GIS-based visualizations and customized software tools to derive actionable insights from agricultural data, particularly to inform water and nutrient management decisions. (poster).

In a maize field at Penn State using Croptix sensor for nitrogen stress detection from leaf spectra.

Prior to joining Croptix, I was working as a postdoctoral research associate in the Department of Geosciences at Pennsylvania State University with Tess Russo in the Hydro Research Group. I collaborated with Joshua Elliott and Ian Foster of Center for Robust Decision-making on Climate and Energy Policy (RDCEP) at the University of Chicago for the U.S. groundwater and climate connection project. I was also working on the simulation of groundwater depletion model for India using physical and economic factors with a particular focus on impacts of policy and possible reform pathways.

Machine learning model development and implementation for water level forecasting in agricultural regions of US (paper and talk)

Future projection of groundwater level changes using integrated machine learning model and climate change scenarios (talk)

I have a PhD in Agricultural and Food Engineering (Hydrology and Water Resources) from Indian Institute of Technology Kharagpur, India, where I worked on hydrological modeling, statistical and machine learning models for water-agricultural systems, simulation-optimization models, water level variability and forecasting, agricultural water management, groundwater-surface water interaction, and conjunctive management practices in irrigated agriculture. publications

When I am not working, I like to get outside and go for a walk with my dog. Over the weekends and evening, one might find me creating artworks with acrylics, resin and vibrant color palettes. Here you will find my original abstract, contemporary and modern paintings.