Planting coral by hand is brutally slow. That is why researchers are building underwater robots they hope can do the job far faster.

On Isla Grande in Colombia’s Rosario archipelago, diving guides have gone out almost every day since 2018 to plant coral seedlings on small concrete disks, known as “cookies”. Lavinia Fiori, one of Isla Grande’s guides and the architect of its restoration project, said the team restored a significant portion of the island’s reef before a mass bleaching event hit in 2023.

“We lost it all. Everything,” Fiori says.

The loss on Isla Grande sits inside a much bigger crisis. From 2024 to 2025, the northern reaches of Australia’s Great Barrier Reef lost nearly a quarter of their coral cover. Between January 2023 and September 2025, extreme heat bleached 84 percent of reefs worldwide.

Coral reefs cover less than one percent of the seafloor, but a quarter of all marine species depend on them for survival. Half a billion people rely on reefs for food and income.

“This is the first time we are at risk of losing an entire branch of the tree of life,” says Alex Neufeld, the science program manager at the Coral Restoration Foundation, a Florida-based conservation organisation. “It would be like losing trees. Not some trees, all trees.”

The standard method of restoration, divers planting coral fragments one at a time by hand, cannot keep up with the losses. The $10 million Mars Coral Reef Restoration effort has been working since 2006 to restore roughly 27 acres of reef habitat globally at a rate of about 440 yards of reef per day.

“The work is slow and takes a lot of labor,” says Edwin de la Rosa, a member of Isla Grande’s nativos, or Afro-Colombian, community and one of its coral restorationists. “We have to go out every day to care for the coral. We spend a lot of time in the water.”

Ian Enochs, the coral program lead at the US National Oceanic and Atmospheric Administration’s Atlantic Oceanographic and Meteorological Laboratory in Florida, put it more bluntly.

The math just “doesn’t math,” he says.

Enochs is among the researchers backing a more automated approach, using robotics, artificial intelligence and autonomous vehicles to scale up reef restoration.

Benjamin Moshirian, a robotics expert with the Australia-based Reef Restoration and Adaptation Program, is leading development of the Deployment Guidance System, or DGS, which he describes as the world’s first automated underwater coral planter.

“You could call it an ocean tractor,” Moshirian says.

Attached to a vessel roughly the size of a fishing boat, the DGS can be piloted by a crew or operate autonomously. A camera paired with artificial intelligence identifies suitable planting sites. When it detects the right seafloor substrate, depth and water flow rate, a dispenser on the back of the craft drops a ceramic plate carrying a coral seedling.

The process takes about half a second, according to Moshirian.

He said each DGS should be able to deploy around a million coral seedlings over its operational lifetime at a cost of roughly US $1 per seedling. He compared that with existing techniques used by the Mars Coral Reef Restoration team, which cost around $8 per seedling.

Moshirian is also working on a “DGS Lite”, a lower-cost version built with GoPros and other consumer hardware. He said he wants a version that smaller projects can afford.

Regulations permitting, he hopes a DGS will one day run night and day with little human input.

“It’ll disperse five juvenile corals per square meter, constantly,” he says. “How else could we hope to achieve that?”

But speed is only part of the problem. Rising ocean temperatures mean restoration efforts also need corals that can survive harsher conditions.

“What’s the point in planting more if they don’t survive?” Neufeld says. “You can seed a million corals and lose them all in a single bleaching event.”

At his Florida lab, Enochs exposes corals to heat to identify specimens that can better handle warmer water. After collecting healthy corals from a reef, his team breaks them into fragments and lets them adjust to tank life. Then the team heats the water to levels that would usually damage the corals.

Some die and some survive. Enochs said the survivors are then planted back into reefs, where they will multiply and produce descendants with a better chance of surviving a mass bleaching event.

Finding those corals “is a needle-in-a-haystack kind of search,” Enochs says. “We’ve got to be evaluating orders of magnitude more coral … than we have in the past.”

He said that kind of work cannot be done by hand, which is why he brought robots into the lab.

Enochs describes himself as “just an ecologist trying to solve a problem.” He said his team already knows stress-hardening can work.

“We know that we can stress-harden corals,” Enochs says. “We know that we can grow corals fast in the lab. We know that we can identify resilient genotypes. We know all of these things. But we’ve got to do a lot of it.”

Melanie Olsen, who leads the ReefWorks program for the Australian Institute of Marine Science, said she sees fleets of robotic systems working together in future, with drones mapping reef health, autonomous submersibles scanning reef structures and surface vessels coordinating operations.

But on Isla Grande, that kind of technology still feels distant. Fiori, who has spent more than a decade working with Isla Grande’s nativos community on reef restoration, said the appeal is obvious.

“Sure, robots would be nice,” she says. “Of course we want help! But where is it?”

Most coral restoration projects around the planet are small, community-based and chronically underfunded.

“I think imagining that just robots will save us could be optimistic to the point of being harmful,” Neufeld says. “There’s never a silver bullet.”

Phanor Montoya-Maya, the reef restoration program manager at the Coral Restoration Foundation, said technology cannot fix the wider threats facing reefs on its own. He said warming oceans, acidification and pollution still need political will and community support.

“If technology is imposed, then we need [community] participation,” Montoya-Maya says. “Restoration succeeds not only because the corals survive, but because of the relationships, the trust, and the stewardship built in the process.”

Read more from Smithsonian Magazine.