In recognition of Earth Day 2020, Intel announced that it is partnering with Accenture and the Sulubaaï Environmental Foundation on a new initiative known as Project: CORAIL. The project aims to monitor, characterize and analyze the resilience of coral reefs using AI or artificial intelligence.
Since first being deployed to the reef around Pangatalan Island in the Philippines last year, Project: CORAIL has collected 40,000 images, which researchers will now use to gauge reef health instantly.
The United Nations Environment Programme (UNEP) considers coral reefs to be endangered. Bottom trawling, overfishing, unsustainable development on the coast, and warming temperatures are considered to be the main causes.
Project: CORAIL will combine the expertise of Intel, Accenture, and the Sulubaaï Environmental Foundation to assist researchers in restoring and supplementing the presently degraded reef.
A key element of Project: CORAIL is a concrete underwater platform known as Sulu-Reef Prothesis, which was designed and built by the Sulubaaï Environmental Foundation. The Sulu-Reef Prosthesis supports unstable coral fragments and manages to incorporate living coral fragments. Since the prosthesis can expand and grow, it can serve as a suitable habitat for marine life.
The Sulu-Reef Prothesis works in conjunction with special underwater video cameras that make use of Accenture’s Applied Intelligence Video Analytics Services Platform (VASP). VASP not only detects and photographs fish, but VASP utilizes AI to count and classify marine life. This data gets sent to a surface dashboard, giving researchers real time trends and analytics. Researchers can use this data to make rapid decisions in order to protect the reef.
The Video Analytics Services Platform utilizes Intel technology, including Xeon processors, Movidius VPU, FPGA Programmable Acceleration Cards, FPGA Programmable Acceleration Cards, and the Distribution of OpenVINO toolkit.
The next generation of Project: CORAIL is expected to use more advanced AI, including an optimised convolutional neural network. As the project evolves, engineers may add infrared cameras to gain a 24/7 picture of the ecosystem, including the migration patterns.
