Dell Technologies is taking a proactive approach to protecting Australia's Great Barrier Reef by analyzing coral reef images in real time using new deep learning technology and edge solutions.

Dell announced a new deep learning model on the 31st in collaboration with Citizens of the Great Barrier Reef in Australia as part of its ESG vision.

This model will be applied to the Great Reef Census (GRC), a citizen-led environmental conservation research survey, to enable citizen scientists around the world to quickly and accurately analyze coral bleaching on the Great Barrier Reef, a World Heritage Site.

Since late 2020, Dell has been deploying edge solutions on ships around the Great Barrier Reef to capture and transmit real-time images of the coral ecosystem. Analysis of these captured images has been performed manually by volunteers, but with the introduction of the latest deep learning model, analysis can now be done automatically.

The time it takes to analyze a single photo is expected to be shortened from the existing 7-8 minutes to less than 1 minute. In the first round of research, it took 1,516 hours to analyze 13,000 images, but the latest deep learning model is expected to quickly analyze the same amount of data in less than 200 hours, improving research results.

Based on this, citizen scientists plan to respond immediately in situations that require rapid recovery or during times of special attention, such as the spawning season.

The deep learning model using the semantic segmentation method used in this study is trained based on Dell's high-performance computing (HPC) graphics processing unit (GPU) acceleration system, and the data is stored on the 'Dell PowerScale' system, a data lake solution for unstructured data.

Additionally, it supports AI learning clusters and AI inference engines through ground computing platforms such as Dell PowerEdge servers.

The project embodies Dell's ESG vision to proactively address climate change and enhance sustainability by working with customers, partners, suppliers, and local communities.

The GRC project is a partnership spanning the Asia Pacific and Japan (APJ) region, bringing together the expertise and technology of Citizens' Coalition for the Great Barrier Reef, Dell Technologies, The University of Queensland (UQ), James Cook University (JCU), Sahaj Software Solutions, and citizen scientists.

Andy Ridley, founder of Citizens' Coalition for the Great Barrier Reef, said, “Dell’s edge solutions and deep learning models are easy to replicate and scale.“It can be applied to coral reef communities around the world,” he said, adding, “The first region outside of Australia where it will be applied is Indonesia.”

Kim Kyung-jin, President and CEO of Dell Technologies Korea, said, "Dell Technologies is expanding technological collaborations with various organizations working to mitigate climate change. Dell's deep learning technology has significantly advanced coral reef research by enabling researchers to quickly access high-quality data. We will continue to expand our vision for sustainability and environmental protection, working with local communities."

Meanwhile, Dell will use its new deep learning model to support the analysis of 42,000 images collected from 314 coral reefs within the 2,300-kilometer coral reef marine park by the end of the year.