EOfactory scales up to provide geospatial users nearly limitless image processing capabilities using CATALYST Microservices Platform.

Singapore-based company EOfactory.ai is an online platform where clients access, process, and extract actionable information from diverse remote sensing data sets. Since then, the Singapore firm has emerged as the world’s ‘go-to’ online production resource for private- and public-sector organizations that need to process large volumes of geospatial data.

With a rapidly growing client base, EOfactory.ai is not dwelling on its success, however. The company is taking dramatic steps to scale its platform capabilities to meet the demands of a new era in Earth observation technology defined by an ever-increasing number of imaging sensors capturing more complex data sets on revisit schedules once thought impossible.

“We built EOfactory.ai to be the world’s first virtual factory for Earth observation. This online factory is now ready to churn out amazing workflows that people can use to build their own products.
These can be simple value-add products, insights, or BIG data analytics. Ultimately, the True AI Earth observation framework platforms will help countries make better decisions around sustainability and food security for future generations.” said EOfactory.ai CEO Abhay Mittal.

EOfactory.ai is a sophisticated yet easy-to-use factory that is designed to enable end-users to derive actionable insights from Earth observation imagery. Starting with data access, the site offers numerous APIs that connect directly with websites operated by remote sensing data vendors, allowing users to query and obtain satellite, aircraft, and drone imagery of any geographic area.

EOfactory.ai offers the option of performing image exploitation to extract geospatial information as a service, or allow customers to do it themselves on a secure EOfactory.ai online workshare. The customer can access a wide variety of AI based algorithms, workflows, and models – or create their own – to process, normalize, enhance, analyze, and generate maps from their selected data sets.

Whether the processing is conducted by EOfactory. ai or an end-user, many of the algorithms and tools running in the background are powered by CATALYST Enterprise (formerly GXL) from PCI Geomatics.
Among the most important of these are the CATALYST Analysis Ready Data (ARD) functions that enable EOfactory.ai users to prepare extremely large image data sets for the application of Artificial Intelligence (AI) and Machine Learning (ML) technologies.

EOfactory.ai has developed AI and ML tools for its customers – with no coding skills required – to use in building custom analytical models for extraction of intelligence from geospatial files. More importantly, EOfactory.ai enables users to set up easy-to-understand visual dashboards to share results with any authorized viewer in any location at any time. Organizations are leveraging these resources to establish production pipelines that continuously analyze new imagery to monitor changing conditions over large geographic areas.

EOfactory.ai sees the need to significantly upgrade the scalability of EOfactory.ai’s processing capabilities in terms of speed and volume as petabytes of imagery, primarily from new remote sensing satellites, become available on more frequent revisit cycles. To accomplish this, EOfactory.ai is now implementing CATALYST Microservices, a new product that optimizes image processing with nearly limitless scalability on the public cloud.

CATALYST Microservices eliminates time-consuming and expensive data transfer by moving the image processing and analysis from local servers to the public cloud where image data is already stored.
Microservices then taps the immense power of cloud computing to cost-effectively complete the project in a fraction of the time required for server processing.

“A nation-wide orthorectification of 10,000 images that now takes three days will be finished in five minutes with CATALYST Microservices, and it opens up the possibility to deliver not only national-level information products but also global datasets,” said Mittal.