Let’s put a couple of numbers into perspective.
There are currently over a thousand EO (Earth Observation) satellites orbiting the earth. Each day, these satellites capture hundreds of terabytes of data. Considering the lower end of the spectrum, if a satellite transmits even 100 TB of data per day, that is approximately 100,000 GB of data. A conservative estimate would suggest that 100,000,000 GB of data is transmitted every day through EO satellites around the earth. And we’re downplaying this number!
The wealth of information that lies in this data has enormous upside potential but without the compute capabilities or the means to extract meaning from all this information, the data is entirely useless. Supporting access to this data is vital with more satellites being launched into space, and platformization is imperative to derive meaningful insights from the information collected.
Platformization is predominantly powered by the cloud. Industry leaders such as AWS, Azure, Google, and others are making significant headway and offering off-premises compute power, allowing people to access services, software, hardware, or even server compute. The platformization of space data has made huge strides in helping create insights that are powering business and organizations around the world.
Let’s consider Earth observational data. Historically, this data was difficult to receive and challenging to use, primarily because it consisted of hundreds of images of the Earth that serve various purposes. However, over the years, the quality and efficacy of the data has improved significantly to be able to provide images of better resolution as well as faster transmission of information which promotes recency of the visuals. What used to be a couple of images a day, has now turned into hundreds of images each day. Processing this volume of information requires compute power that Azure, AWS and the likes can provide, helping host and serve this data to transform into meaningful insights.
Today, the space industry is undergoing an analytical paradigm shift. PaaS providers such as AWS are not only looking at helping access the data, but are also creating a layer of analytics on top of it. Recently, AWS announced Geospatial ML with Amazon Sagemaker which enables teams to build, train, and deploy ML models faster using geospatial data. This is a huge leap forward considering the access to multiple geospatial data sources, pretrained ML models, and built-in visualization tools to run geospatial ML faster and at scale.
Initiatives such as these allow for a wide range of use cases that are led by satellites and powered by data-driven analytics. For instance, the data could now be looking at visuals of the Earth and providing nuanced information such as the vegetation, the quality of the soil, the health of the vegetation, what kind of crop yield is available and other insights that would have previously required physical presence and testing.
However, the most important part remains connecting these analytics to the downstream systems or operational systems which are already in place. Integrations are essential to use the data in an actional format across the system of record or system of insight the user currently operates on. This reduces the dependency of copying and pasting data from one system to another and instead facilitates real-time or near real-time assimilation of data and insights.
While AWS and Azure have provided platformization, newer and more specialized entrants in the satellite space have been making strides as well in transforming hundreds of TB in satellite data generated each day. Organizations like Airbus and Maxar not only provide data from satellites but also deliver analytics as an added layer of service on top of the data. Similarly, BlackSky, offers insights on the imagery generated such as how many aircrafts might be present in an airfield or measuring an oil spill that might occur using data generated from the satellite. The recent acquisition of Sinergise, a developer platform for EO data, by Planet Labs also showcases the commitment of newer players allowing customers to access multi-source EO data for processing, analysis, and insight extraction. Other solution providers such as Cyient have invested decades in the industry and help organizations create visualizations and dashboards, delivering trends from data patterns to take meaningful actions.
Specialists can now tune the data using a convergence of IT/OT and data science to solve business problems. As organizations move towards platformization, customers are becoming more empowered to make some of these decisions themselves. Digital transformation enables organizations to move away from being services-driven to being solution-driven, where a satellite provider who might now understand the mining industry in the past is able to see the data and drill down into the core of the business problem.
Simplified. It’s what platform providers like AWS and Azure have done to the vast majority of people. They have simplified the way data is ingested, transformed, and delivered so customers can make meaning of them without the burden of having data science specialists break down the minutia of details within them. With satellite data increasing in volume, velocity, and veracity, acting quickly could mean the difference between a couple of million dollars or even saving lives. This is what makes platformization the backbone, holding up space as one of the megatrends of the decade.
About the Author
Gareth has been with Cyient for 17 years in various roles, most recently as Head of Location Based Solutions. He has had a passion for geography since high school, where his teacher told him to get into an emerging technology called "GIS." Since then, he has maintained his passion for GIS and Geospatial by working closely with our largest Geospatial customers as well as industry startups and SMEs.
Let Us Know What You Thought about this Post.
Put your Comment Below.