The Australian Institute of Marine Science (AIMS) deployed a fleet of autonomous underwater vehicles during this year’s coral spawning event on the Great Barrier Reef, collecting high-resolution data on reproduction patterns across a much larger area than previous studies.

Coral spawning happens once a year, typically in November or December, when corals release eggs and sperm simultaneously in a spectacular underwater snowstorm. The timing and success of spawning are critical indicators of reef health, but monitoring has traditionally been limited by the number of human divers you can deploy.

The autonomous vehicles, developed in partnership with the University of Sydney’s robotics lab, can operate for up to 72 hours without human intervention. They use computer vision algorithms to identify and count coral colonies, detect spawning activity, and collect water samples for later analysis.

Dr. David Wachenfeld, AIMS’s chief scientist, said the technology lets researchers study spawning at a scale that wasn’t previously possible. “We can now monitor dozens of sites simultaneously and capture data throughout the night when spawning occurs. That’s fundamentally changed what questions we can ask.”

This year’s spawning was particularly important because it followed a mild summer with no major bleaching events. Reef researchers are cautiously optimistic that several consecutive years without severe bleaching could allow coral populations to recover somewhat.

The robots collected data on spawning synchrony, which is when different coral colonies release gametes at the same time. High synchrony improves fertilisation success, but recent studies suggest climate stress may be disrupting the environmental cues that coordinate spawning.

That’s concerning because poor fertilisation means fewer coral larvae, which means slower reef recovery. Understanding the mechanisms behind spawning coordination could help predict how reefs will respond to continued warming.

The autonomous monitoring system uses machine learning models trained on thousands of hours of underwater video. The models can distinguish between different coral species, identify spawning events, and even estimate the volume of gametes released.

That last capability is particularly clever. The robots illuminate the water with structured light patterns and use the distortion of those patterns to estimate particle density. It’s similar to techniques used in industrial quality control, adapted for underwater use.

The research builds on earlier work using fixed underwater cameras to monitor specific reef sites. Those cameras generated valuable long-term datasets but couldn’t cover large areas. The mobile robots fill that gap.

AIMS is also using the robots for other monitoring tasks, including tracking crown-of-thorns starfish populations and surveying cyclone damage. The platform is flexible enough to carry different sensor packages depending on the research question.

One challenge with autonomous marine robots is communication. Radio signals don’t propagate well underwater, so the vehicles rely on acoustic modems for short-range communication and surface periodically to upload data via satellite.

The robotics team is working on better autonomous navigation algorithms that would let the vehicles operate in complex reef environments without getting tangled in coral or crashing into underwater structures. That’s harder than it sounds because underwater visibility is often poor and GPS doesn’t work underwater.

The spawning monitoring project received funding from the Reef Trust, which supports research and management of the Great Barrier Reef. Total project cost was around $4.5 million, covering vehicle development, sensor systems, and three years of field operations.

Some environmental groups have criticised the focus on monitoring technology, arguing that the money would be better spent on direct interventions like coral restoration or water quality improvements. AIMS counters that you can’t manage what you don’t measure.

That debate reflects a broader tension in reef management. The Great Barrier Reef faces multiple stressors including warming ocean temperatures, ocean acidification, water quality issues from coastal runoff, and crown-of-thorns starfish outbreaks. Addressing all of those simultaneously requires both better monitoring and active interventions.

The autonomous monitoring data will feed into the Reef 2050 Plan, which guides management decisions about the Great Barrier Reef. That plan is updated every five years based on the latest scientific evidence.

Whether better monitoring leads to better management outcomes depends partly on political will and partly on what the data reveals. If coral spawning is declining despite mild conditions, that suggests the reef’s resilience is eroding in ways that are difficult to reverse.

The robots will continue operating throughout the year, though spawning season is their primary mission. AIMS plans to expand the fleet to 20 vehicles by 2027, enabling near-continuous monitoring of key reef sites.