The first images of Earth from space were taken in 1946 from aboard a V-2 rocket sent to space by the United States. They had no market as such, but less than two decades later the first weather satellites sent their images back to Earth. Ever since, the amount of satellites, of the data they capture, and the applications and business models they serve has vastly increased.
To be able to make use of this data and to eventually turn it into insights requires solutions to store, process, and analyse data and to make the data itself as well as the retrieved information and associated services accessible and usable. This asks for online, Cloud-based solutions with high performance computing and innovative processing and analytical capabilities.
And such solutions exist plenty, provided by ICT players, Earth Observation (EO) companies, as well as a growing number of start-ups and SMEs with solutions for different aspects that can be divided into four major tiers:
- The “Data Generation Tier”, which includes open EO data catalogues (e.g. Sentinels, Landsat), proprietary data from private operators, or higher-level/derived products.
- The “Resources Tier”, where Cloud-based computational services enable users to process data in the Cloud.
- The “Platform Tier”, which provides web-interface services and toolboxes to add value to the data and produce or expose services.
- The “Knowledge Tier”, where end-users gain access to the derived insights.
Navigating the complex landscape of EO platforms and understanding what each platform has to offer is essential for companies and entrepreneurs who want to bring their services to the market. However, there is currently a lack of consolidated knowledge about the existing resources – datasets, platforms, communities – and their fit with a specific task. In particular, an overview of sources of in-situ observations seems unavailable. And even if platforms have been identified, they vary to a degree that making a choice is not always obvious: the type of data they offer (e.g. global or local, real-time, spatial, from different sources), the quality of the data, the price of the offering, data policies, data that is differently curated and archived, different APIs to interact with thematic data etc., and certain tasks may require using multiple platforms.
At a high-level perspective related to platforms in different or even across tiers, the following platforms should be mentioned:
- The Copernicus Data and Information Access Services (DIAS) aim to ensure easy, cost-efficient and reliable access to Copernicus data and provide storage, computing, processing tools, software packages, etc. Today, five such services are available: Creodias, Mundi, ONDA, Sobloo, and Wekeo, each with a specific value proposition on top of basic access to data. A good overview on the capabilities of each DIAS is being maintained by EARSC.
- Google Earth Engine and Amazon Web Services are the most widely used platforms. They offer access to major open-source EO data catalogues alongside scalable Cloud-based infrastructure.
- Satellite imagery providers are increasingly shifting to data analytics services executed through their own platforms. Examples would be Maxar’s GBDX and Planet’s Explorer.
- Sector-specific platforms providing fit-for-purpose analytical tools are also on the rise. Thus, a number of Thematic Exploitation Platforms have been launched, alongside FAO’s SEPAL and many other web portals.
But there is much more out there. Within e-shape’s Work Package 3, a high-level framework is being developed to understand platform offerings and their APIs, and to catalogue these platforms. More than 30 such platforms, mostly located in Europe, have been identified and classified. With further feedback from e-shape pilots, the framework will take shape and will be made available to contribute to the European GEO System of Systems. The e-shape Market Trends Observatory will be closely monitoring the progress of this work and all associated developments in order to produce insights that could be used by e-shape pilot partners and the EO community at large.
This article has been informed by Marie-Francoise Voidrot-Martinez from the Open Geospatial Consortium.