AI helps digital twin scale coral reef restoration

Underwater nursery image with possible coral growths on metal poles and table structures segmented with purple highlight and red outline. — *Segmented images of underwater coral nursery structures illustrate potential for applications of AI*

Tech experts are working with a leading research university to develop a first-of-its-kind environmental AI solution for a digital twin that can help predict how coral reefs will behave during restoration projects.

UK-based global leader in artificial intelligence (AI) digiLab has teamed up with the King Abdullah University of Science and Technology (KAUST) in Saudi Arabia to develop and deliver an AI solution that will be used to predict and protect the success of a 100-hectare reefscape initiative in the Red Sea. This pioneering work forms part of the world’s largest coral restoration project, the KAUST Coral Restoration Initiative (KCRI).

AI integration is being led by digiLab co-founder Dr Anhad Sandhu, with support from Dr Liz Goergen, Head of Monitoring Visualisations and Database Management at KCRI. This two-year programme of international collaboration will use AI to fast track reef restoration and conservation, at scale, for the first time.

Understand and enhance reef resilience

The AI solutions tackle critical aspects of coral resilience, such as thermal tolerance and growth. They can revolutionise how resilience is understood and enhanced in response to changing environmental conditions.

Deploying AI tools enhances the capabilities of environmental sensors across the reefscape. AI solutions not only reduce the time required for coral assessments but also increase the accuracy of data retrieved from key environmental stressors like temperature and light, crucial for propagation and survival of coral species.

Setting a new standard for coral restoration schemes, the integration of AI demonstrates a scalable model for environmental conservation. The project offers insights and methodologies that can be adapted globally to protect and restore vital reef ecosystems in the face of continued environmental challenges.

Coral nursery system to plant millions

The KCRI project is a global demonstration of coral restoration that integrates international expertise with pioneering propagation, planting and monitoring approaches to accelerate solutions for reef ecosystems.

The project focuses on restoring and enhancing reefs through a large-scale coral nursery system to support propagation and natural reproduction. With KAUST planning to plant two million corals around the reefscape by 2030, the nurseries must be efficient. So, at peak capacity, the sea-based nurseries will target production of 100,000 corals a year, with the primary land-based nursery aiming for as many as 400,000.

Using advanced AI, project scientists can now closely simulate and predict behaviour of the real-world reefscape in real-time. AI tools delve into the ecosystem to uncover crucial data, such as algae growth and signs of disease. This allows for easy and accessible monitoring in nurseries both on land and in the sea.

Large language model and digital twin

Throughout this project, digiLab will assist KAUST in three key areas:

Coral Measurement: digiLab is assisting KAUST with computer vision by developing an analysis pipeline to enable more efficient video monitoring. This aims to reduce monitoring time of corals from two months to just two weeks, enabling faster and more comprehensive tracking.
Optimal Sensor Placement: digiLab’s innovative sensor placement system augments data collection, using machine learning to optimise deployment. This enhances efficiency and cuts costs by reducing the number of sensors and total time needed for data collection, whilst maintaining accuracy in monitoring vital statistics, including temperature, dissolved oxygen, photosynthetic active radiation, wind and tide.
chatReef: The full digital twin is projected to reach 22 petabytes (22 million gigabytes) in size, making data-driven decision-making challenging. So, to provide an intelligent way of searching the data within the system and maximising its value, digiLab is integrating its agentic large language model (LLM) platform to interact with the vast unstructured and structured data held within the digital twin.

Together, the project partners are blazing a trail for AI, says digiLab Co-founder, Dr Anhad Sandhu:

“Our partnership with KAUST on the largest coral restoration project marks a significant milestone in the application of AI technology to solve the world’s challenges. By developing AI models for the reefscape’s digital twin, digiLab is supporting revolutionary methods for coral-reef monitoring and restoration.”

The project will boost effective restoration of reefs globally, adds KCRI Director Professor Dave Suggett:

“By partnering with digiLab to develop an AI-enabled digital twin of the reefscape, KCRI is pioneering innovative approaches to upscale coral management efforts, including restoration decision-making.

“This collaboration allows us to leverage cutting-edge AI technology to monitor, predict, and optimise our restoration strategies in real-time, potentially revolutionising how we approach coral reef conservation on a global scale.”

Innovation is key on a project of this size, with learnings for the wider coral reef and restoration community, concludes Dr Liz Goergen, Head of Monitoring, Visualisations and Database Management at KCRI:

“Coral reef restoration has never been done at the scale we are planning. Therefore, to accurately and efficiently plan, monitor and measure progress and success, we needed to look outside our typical toolbox.”

Maths and data, science and tech

Having grown from a small team of expert mathematicians and data scientists, digiLab is now a leader in uncertainty quantification, pioneering the use of AI in highly regulated or safety-critical industries.

Based on years of cutting-edge academic research, the digiLab team teaches clients how to solve complex engineering, infrastructure or data challenges using the company’s no-code agentic AI platform.

Established in 2009, King Abdullah University of Science and Technology (KAUST) is an interdisciplinary graduate research university based in Saudi Arabia, with more than 120 different nationalities on campus.

Working on some of the most pressing scientific and technological challenges in the world, KAUST specialises in areas of food and health, water, energy, environment and the digital domain.

Further Reading: