Robots Are Planting Coral Faster Than Humans Ever Could

Robotic coral restoration is the use of autonomous underwater vehicles and AI guided robotic arms to seed and plant coral at a scale human divers cannot match, and it matters because coral reefs shelter a quarter of all marine species while supporting the livelihoods of hundreds of millions of people, yet the world has lost a large share of its coral to warming seas. The Great Barrier Reef alone has suffered repeated mass bleaching events since 2016. Traditional restoration, done by hand at the pace of divers working brief bottom times, cannot keep up. Robots might.

Two projects, one delivering coral babies and one planting coral on manufactured skeletons, show how automation and AI are changing the mathematics of reef repair in Australia.

LarvalBot, the Undersea Seeder

LarvalBot is a briefcase sized submersible robot developed by Queensland University of Technology researcher Matthew Dunbabin, building on his earlier RangerBot platform for monitoring the Great Barrier Reef. In late 2018 it made the world's first robot delivered settlement of baby coral on the reef. Working with coral scientist Peter Harrison, whose research showed that coral larvae collected during mass spawning events can be settled onto damaged reef areas, the team loaded the robot with hundreds of thousands of microscopic larvae and released them precisely where damaged coral needed recolonizing.

The robot's capacity per mission was initially around 100,000 larvae, with plans shared by the team to scale to millions per run. The Great Barrier Reef Foundation, which supported the work, described the ambition plainly: get larvae onto damaged reefs far faster and more precisely than teams of divers. The robot navigates autonomously along the reef, avoiding obstacles using its vision systems, and gently disperses larvae that over weeks settle and develop into coral polyps.

Coral Maker Borrows From Factories

On the other side of Australia, marine biologist Taryn Foster founded Coral Maker with a different insight: the bottleneck in coral propagation is manual labor, and industry has already solved mass production. Coral Maker adapts technology from the masonry manufacturing industry, using molds designed in Autodesk Fusion 360 to mass produce limestone coral skeleton plugs, then seeding them with coral fragments. The seeded plugs are planted on reefs, giving corals a head start on structure that would otherwise take years to accrete.

The next step is robotic. With pro bono support and AI tools from Autodesk and the Autodesk Foundation, and collaborative robots from Universal Robots, Coral Maker is automating the seeding process. AI vision systems identify coral fragment size and orientation so robotic grippers can handle living tissue without damaging it. The company has piloted its system off the coast of Western Australia near the Abrolhos Islands, growing first trial batches of coral on its manufactured skeletons for over a year. Its stated goal is restoring hundreds of hectares of reef per year, a scale that manual methods cannot approach.

Why Speed and Scale Are the Whole Game

A human diver can plant a few hundred coral pieces in a day. Reefs are measured in thousands of square kilometers. Mass coral spawning produces billions of larvae a few nights a year, a resource that currently washes away largely unused. The logic of both projects is to attack the ratio, borrowing speed from robotics and precision from AI so that restoration can operate on something closer to the scale of the damage.

Neither project claims robots alone will save reefs. Foster has been clear that climate change remains the underlying threat and that restoration buys time, not immunity. Heat waves can kill planted coral as readily as wild coral. But healthy restored patches can seed surrounding reef, and emerging research on heat tolerant coral strains gives restoration efforts a harder biological foundation to build on.

From Pilot to Industry

Coral Maker is now seeking investment to scale from pilot to operation, and reef restoration more broadly is professionalizing, with the Australian government funding restoration and adaptation programs and similar efforts spreading from the Caribbean to Southeast Asia. The combination of factory methods, autonomous robots, and computer vision hints at a future where reef repair is an industry rather than a volunteer activity.

On the night of a mass spawning, when corals release clouds of larvae under the moonlight, most of that potential is lost to the currents. Catching even a fraction of it, and placing it exactly where a reef is broken, is a job tailor made for patient machines. The robots are ready for the night shift.