Deep Dive: How IoT and AI May Help Save Coral Reefs

The oceans are vast, mysterious, and home to millions of species of marine life, some beautiful and others equally devastating. Researchers are using cameras, artificial intelligence, and machine learning systems to capture data about coral reefs and ensure they remain healthy and diverse marine ecosystems.

 

Article Key

Image credit: Queensland University of Technology 

The Earth’s coral reefs are a perfect combination of beauty and purpose. They provide habitats and shelter for hundreds of marine life species and generate billions in tourism, but they also protect coastlines against storm damage and provide food and income for millions of people.

Coral reefs have been diminishing due to several factors, including climate change, human sabotage, overfishing, and coastal development. Divers typically monitor coral reefs by taking images and counting fish while underwater, but that can be slow, dangerous, and disruptive. In recent years, tech companies have been joining forces to help conservationists protect coral reefs.

Off the CORaiLs

The degradation of coral reefs led the Sulubaaï Environmental Foundation, a conservation organization based in the Philippines, to team up with tech leaders Accenture and Intel. The group launched Project: CORaiL, which is an edge-computing solution that uses artificial intelligence to assess the coral reef near the Pangatalan Island on the western side of the Philippine Islands.

A key mission of Project: CORaiL is to monitor the health of the reef by identifying and tracking the diversity and abundance of fish that inhabit it. Engineers designed a man-made concrete structure, called a Sulu-Reef Prosthesis, that will support coral growth to rebuild and stabilize the reef. The prosthesis has integrated video cameras that are equipped with the Accenture Applied Intelligence Video Analytics Services Platform (VASP).

Powered by Intel® Xeon® processors and integrated with Intel FPGA Programmable Acceleration Cards and Intel® Movidius Vision Processing Units, the VASP can detect and photograph marine life as they swim past. Machine learning algorithms classify and count the fish without disturbing the environment. The collected data is sent to a land-based system, providing streaming video and real-time analysis that researchers can study to analyze fish populations to determine reef health.

A second iteration of Project: CORaiL is in development and will include an optimized convolutional neural network; researchers are also considering the integration of infrared cameras to enable night video capture. In addition to gauging the health of coral reefs, the Project: CORaiL technology could be used to study migratory patterns of fish and intrusion in protected waters.

Yellow bot used in coral reefs

Image credit: Queensland University of Technology 

A Thorny Issue in the Reef

The Great Barrier Reef off the coast of Queensland, Australia has been deteriorating as crown-of-thorns starfish (COTS) eat away at the living structure. More than half of the Great Barrier Reef has been destroyed by these spiky invertebrates. Australian and Iranian researchers aim to reduce their destructive impact through the use of autonomous underwater vehicles that communicate with each other.

The small, torpedo-shaped robots swim around the reef to monitor and control the presence of the damaging starfish. The COTSbots, developed at the Queensland University of Technology, are equipped with cameras, a computer vision system, and a robotic arm with a needle at its end. Using a GPS navigation system it collects images, which are analyzed by a deep neural network trained to identify COTS. When the underwater drone confirms the presence of the coral-eating creature, the robotic arm extends and injects it with bile, which eventually kills the offending sea creature.

A newer version, called the RangerBot, is in development. It uses two stereo cameras systems and computer vision to navigate underwater and detect and destroy the COTS. The goal is to make them smaller, more maneuverable, and less expensive so researchers can deploy a fleet of them to monitor and protect coral reefs.

 

Intel, the Intel logo, and other Intel marks are trademarks of Intel Corporation or its subsidiaries.

Find Intel® IoT Market Ready Solutions

Search the Intel® IoT partner ecosystem and discover 130+ end-to-end IoT solutions.

Become a Member of the Intel® Partner Alliance

If you are delivering repeatable and scalable IoT offerings using Intel technology, the Intel® Partner Alliance program gives you more opportunities to build your business. Get started.