The Australian Geospatial-Intelligence Organisation’s (AGO) Analytics Lab Program (AGO Labs), coordinated through FrontierSI, was created to trial innovative ways to collaborate and work with industry. The program aims to address AGO challenges and build new industry capability through a selected short-term project with Australian companies. The first round of AGO Labs focused on machine learning and analytics challenges for producing automated imagery analysis.

FrontierSI chief executive officer Graeme Kernich announced the three successful partners for the 2020 program.

“We’re excited to announce that the successful partners in the AGO Labs Program are Australian Earth Observation specialists Ozius, Australian policy strategists Urbis and the Australian Institute of Machine Learning (AIML), and US space infrastructure company Maxar Technologies working with its Australian partner Geospatial Intelligence.”

‘’This exciting opportunity assists AGO to extend its capability and access those innovation companies to trial new technology approaches. At the same time, the program enables businesses to showcase their skillsets to AGO,’’ Mr Kernich said.

‘’FrontierSI has been integral in the development of the program and we believe AGO Labs is a great mechanism to expose new specialist capabilities to agencies such as AGO in an innovative way.

‘’From this process, learnings can be gathered about the barriers and opportunities for AGO to work with an increased breadth of companies. In turn, it could lead to a long-term mechanism for AGO to access and provide a pipeline of activities to industry for testing new innovations and thinking in analytics,’’ he said.

AGO acquires vast amounts of commercial satellite imagery data over much of the world in order to produce geospatial information and intelligence for its Defence and Government customers. The data collected by AGO are notable for high-resolution, high-revisit rates, and are typically exploited using largely manual/supervised processes in order to create a variety of products.

For the 2020 program, the AGO nominated novel capability challenge topics including ‘beautiful contours’, ‘features unsupervised’ and ‘low-cost object identification models’:

• ‘Beautiful Contours’ creates an opportunity to explore more advanced ways to meet elevation data needs of Australia’s Defence and broader Commonwealth government agencies;
• ‘Features Unsupervised’ acquires vast amounts of commercial satellite imagery data over much of the world in order to produce geospatial information and intelligence;
• ‘Low-cost object identification models’ explores the use of automated object identification capabilities to complement its existing imagery exploitation processes.

Mr Kernich also noted: “We’re proud to be lead coordinators of programs like AGO Labs, which partner with our government organisations and which will continue to deliver ongoing investment in our space-enabled industries.”