Geospatial data is best suited for machine learning applications. – Soumya Kanta Dash

Interview with Soumya Kanta Dash

Soumya Kanta Dash

Geospatial Data Scientist, GeoAI
New Delhi, India

Q1. Please share your educational and professional journey with us?

My journey started from Central University of Jharkhand (CUJ) Ranchi where I completed my Integrated M.Tech in Geoinformatics of which, one year I spent as a Visiting Student Researcher at Advanced modeling and simulation laboratory, Indian Institute of Technology Kharagpur (IIT KGP).

During my coursework I got accepted as an Intern at various prestigious organizations as Indian Meteorological Department and Indian Institute of Technology Kharagpur, where I got into machine learning and AI space to solve problems and develop solutions to real world problems. During this I came across various hackathons, which fueled my passion.

I won various hackathons focusing on geospatial data science organized by various industry leaders and organizations such as Google Earth Engine India challenge 2018, Bangalore Tech Summit Global hackathon 2017, NextEra Energy Satellite Hack 2018.

Currently I am working as a Geospatial Data Scientist at Esri R&D Center.

Q2. What did attract you towards Geospatial Data Science?

During my coursework and research, I observed that AI and machine learning combined with geospatial technology could solve various unanswered questions widely searched by the industry and which could lead to efficiency in previous traditional workflows. I realized that Geospatial data is best suited for machine learning applications due to huge availability of data and capability to provide answers in complicated scenarios.

Q3. You are doing research on GeoAI. Please share some brief about it?

GeoAI, also called “Geospatial artificial intelligence”, is basically the combination of geographical information system (GIS)/Geographic dimension and artificial intelligence (AI). Previously, I worked on various research and building solutions such as Extracting building footprints from satellite imagery, predicting weather at any location using ML, Identification of plants from UAV, detection and geo-location of electric utility features etc. Currently, I am working on use of Image-to-Image translation deep learning models for translating between different domain of imageries or modalities.  Here are links to some of my research:

Urban 3D Structure Reconstruction Through a Generative Adversarial Network Model –

Urban surface simulation through image-to-image translation deep learning algorithm using optical aerial imagery

Q4. What exactly is Geospatial Data and where it is used in the real-world fields?

Any data which has any kind of geographic information associated with it can be called as geospatial data, that can a zip code, location co-ordinates, address or city and off course satellite imagery. About 80% of data generated today comes with some sort of location information. Geospatial data are best kind of data that give huge insights about our real world may it be fastest route to any place, best locations to setup a business, predicting crimes before they happen, predicting weather, cyclone tracks and many more, possibilities are endless.

Q5. You have won many Hackathon. How it helps to compete in real world?

I have been able to win many hackathons organized by industry leaders i.e. Google, NextEra Energy, KRSAC Bangalore, focusing on to leverage geospatial data science for respective fields. Hackathons are best way to solve a challenge a industry is facing in fastest and efficient way possible. Competing in hackathons, boosted by skills in building things and coming-up with innovative solutions in a constricted timespan, which are the two things the industry searches for in a candidate. These competences helped me a lot. 

Soumya Kanta Dash

Q6. How do you see the role of AI/DL/ML and Data Science in the field of Geospatial and GIS/Remote Sensing?

The intersection of AI and GIS is creating massive opportunities that were not possible before. AI, machine learning, and deep learning are helping us make the world better for helping, for example, to increase crop yield through precision agriculture, understand crime patterns, and predict when the next big storm will hit and being better equipped to handle it. AI/DL/ML have opened ways to utilize these huge amounts of data collected every day and gain valuable insights.

Q7. Which one thing do you want to change in yourself and why?

I would favor not to change anything in myself. As nobody is perfect in this world and everyone has uniqueness and flaws. It is this that makes us who we really are. So, I would expect everybody to stay as they are and perform to best of their abilities.

Q8. What are your research interests?

My research interests are currently application of the state-of-the-art deep learning models for various geospatial customer problems, integration and deployment of the models into Esri suite of software’s for all our customers.

Read more about Soumya Kanta Dash @

Facebook –

LinkedIn –