Earth Day 2020: Using artificial intelligence to save coral reefs

AI is used to count and classify marine life and the data is sent to a surface dashboard to track analytics and trends in real time without human interference

AI is used to count and classify marine life and the data is sent to a surface dashboard to track analytics and trends in real time without human interference (Intel)

To mark Earth Day 2020, Accenture, Intel and the Sulubaaï Environmental Foundation have announced Project: CORaiL. Project: CORaiL is an artificial intelligence (AI)-powered solution to monitor, characterise and analyse coral reef resilience. Project: CORaiL was deployed in May 2019 to the reef surrounding the Pangatalan Island in the Philippines and has collected about 40,000 images, which have been used by researchers to gauge reef health in real time.

Coral reefs are one of the world's most diverse ecosystems, with more than 800 species of corals providing habitat and shelter for approximately 25% of global marine life. Coral reefs are also extremely beneficial to humans as they protect coastlines from tropical storms, provide food and income for almost one billion people, and generate $9.6 billion in tourism and recreation each year.

(Intel)

(Intel)

But according to the United Nations Environment Programme, coral reefs are endangered and rapidly degrading due to overfishing, bottom trawling, warming temperatures and unsustainable coastal development.

The abundance and diversity of fish serve as an important indicator of overall reef health. Traditional coral reef monitoring efforts involve human divers either directly collecting data underwater or manually capturing video footage and photos of the reef to be analysed later.

Those methods are widely trusted and employed, but they come with disadvantages. For example, divers can interfere with wildlife behaviour and unintentionally affect survey results, and time underwater is limited as divers can often only take photos and video for around 30 minutes at a stretch.

Engineers from Accenture, Sulubaaï and Intel have combined their expertise for Project: CORaiL with the goal of helping researchers restore and supplement the existing degraded reef in the Philippines. they have built a Sulu-Reef Prosthesis, a concrete underwater platform designed by Sulubaaï to provide strong support for unstable coral fragments.

The Sulu-Reef Prosthesis incorporates fragments of living coral within it that will grow and expand, providing a hybrid habitat for fish and marine life. Then, they have placed intelligent underwater video cameras, equipped with Accenture's Applied Intelligence Video Analytics Services Platform (VASP) to detect and photograph fish as they pass.

(Intel)

(Intel)

VASP uses AI to count and classify marine life without human interference and the data is sent to a surface dashboard. This dashboard provides analytics and trends to researchers in real time, enabling them to make data-driven decisions to protect the coral reef.

"The value of your data depends on how quickly you can glean insights to make decisions from it," said Athina Kanioura, Accenture's chief analytics officer and Accenture Applied Intelligence lead.

"With the ability to do real-time analysis on streaming video, VASP enables us to tap into a rich data source — in effect doing 'hands on' monitoring without disrupting the underwater environment," Kanioura added.

Accenture's VASP solution is powered by Intel Xeon processors, Intel FPGA Programmable Acceleration Cards, Intel Movidius VPU and the Intel Distribution of OpenVINO toolkit.

Engineers are currently working on the next-generation Project: CORaiL prototype, which will include an optimized convolutional neural network and a backup power supply. They are also considering the addition of infrared cameras to the system, which enable nighttime video capture to create a complete picture of the coral ecosystem.

Additional uses of a set-up like this could include studying the migration rate of tropical fish to colder waters and monitoring intrusion in protected or restricted underwater areas.