The origins of SpaceSense: Interview with Jyotsna Budideti

mins watch
mins read
February 28, 2022

SpaceSense was created by Jyotsna Budideti and Sami Yacoubi in 2019. A lot of things have changed since then, but the vision and ambition remains intact: Democratize the use of satellite imagery. In this interview, co-founder and CEO Jyotsna Budideti goes over the motivations that led to SpaceSense’s creation, the current objectives and what you can expect in the coming months.

What was the idea that pushed you to start SpaceSense??

Coming from a machine learning background, when I was first experimenting with satellite data, even the most simple step felt like a huge uphill task to learn what to do & how to use this data. The tools that I took for granted in computer vision & other machine learning problems did not exist and it felt wrong that it should be so difficult to use this technology.
Furthermore,I learned about how this technology can benefit so many industries and society in general. Which made it even more surprising to see that this technology is still not used everywhere by everyone.  Around that time, with the acceleration of the New Space movement, we saw hundreds of satellites being launched in orbit every year with the sole purpose of monitoring every inch of Earth. When you combine that with the even more rapidly advancing AI technology, I could not resist thinking how many more use cases would open up for businesses & organizations to solve their everyday problems. But to achieve that in practicality, to build solutions that can be operational, we have to make Earth Observation data (EO) & AI work together smoothly.
In the Machine Learning (ML) community, tools like scikit- learn, Keras, ResNet,, and the like, democratized ML technology. And they transformed a process that previously required Ph. D level skills into something that a high school student can get started with.
My idea was to do the same for satellite imagery. To make this technology mainstream by enabling AI with EO data, and democratize it by making it insanely simple & fast to build solutions with.
So I started SpaceSense with the goal to become the first satellite imagery technology enabler, that will make it possible for anyone to build solutions with Satellite imagery & AI.

What are the missions of SpaceSense? 

The mission of SpaceSense has always been to enable every data scientist to build meaningful solutions using satellite imagery.
We want to remove the barrier of entry to satellite data, enabling generalist data scientists to use satellite data by removing the need for specialized expertise in remote sensing.  They should get started and build their first application in a matter of hours.
And for the existing users of the satellite data, our mission is to accelerate the process & reduce the operational costs of building their solutions by at least a factor of 5x directly by the use of SpaceSense tools.

Sentinel 5P from the European Space Agency

If we push further and ask why we have this mission, it’s mainly due to climate change. By democratizing satellite imagery technology, we give more power to civilians & government agencies to advance the way we monitor, manage & adapt to climate change and build solutions for sustainable development. Today we have a few thousands of climate scientists trying to solve the world’s million problems. By making the technology more accessible, we enable millions of data scientists to tackle these problems, giving ourselves a much better shot at solving climate change.

What are the current goals of SpaceSense?

Today the biggest challenge for a data scientist in using satellite imagery is in bringing this data to a format they can understand, analyze and manipulate to be ingested by any ML model. This data preparation process with satellite data requires expertise in geospatial analysis today. And so, our current mission is to remove the biggest barrier and bottleneck to satellite imagery by simplifying and accelerating Data preparation for AI. This way, beginners can prepare their data and more advanced users won’t have to worry about wasting a lot of time in pre-processing the data, and they will be able to utilize this time in more valuable tasks.

How do you think SpaceSense is playing or will play a crucial role in the New Space industry?

Every deep technology has an adoption curve before everyone can use it. Like in the case of AI / ML, first it is restricted to the researchers and now, anyone can create their own ML solutions. I believe SpaceSense will be a key player in pushing technology adoption, making satellite imagery mainstream, opening up the Space industry to impact many more industries and decisions across the globe.

WorldView 1 from Maxxar

Where do you see SpaceSense in the next 5 years?

We want to be at the forefront of satellite imagery democratization and help thousands of companies create meaningful insights from satellite imagery without needing to build a whole remote sensing team. We also want to be one of the significant players when it comes to data and process standardization, something that is critically lacking today and contributing to this slow adoption. In 5 years, having satellite solutions should be the norm, not the exception. And SpaceSense will be spearheading that change.

What new solution can we expect from SpaceSense in the near future?

Following our immediate mission to simplify & accelerate data preparation for AI, we are eager to release our product, the complete data preparation toolbox for AI with Satellite imagery very soon. In the coming months the first version of the toolbox will enable easy access to multispectral and radar satellite data from the Sentinel-2 and Sentinel-1 missions with S2 ARD and S1 ARD respectively in a format that is customized to your use case and ready for use by your data scientist. The Data fusion module will enable you to create a rich datacube with information from different sources including different types of satellite data, your own drone imagery, weather data, valuable ground information from your sensors or surveys, and much more so that your data scientist can build the best ML solution possible. 

Want to benefit from geospatial
AI easily?

Connect with us

Check out our upcoming solutions that you can test

Check it out