Scientists have developed an autonomous underwater robot that can locate and map biodiversity hotspots on coral reefs by combining sound and vision in real time, offering a new way to study and protect fragile marine ecosystems.

The system, called CUREE (Curious Underwater Robot for Ecosystem Exploration), was developed under the Woods Hole Oceanographic Institution Reef Solutions Initiative. It uses cameras, hydrophones, and onboard computing to analyze underwater environments while it moves, identifying areas of intense biological activity with high precision.

Unlike traditional reef surveys that rely on divers, the robot operates autonomously for long durations and can cover larger areas without risking human safety. It continuously collects both visual and acoustic data to understand how marine life is distributed across complex reef structures.

Coral reefs occupy less than 0.01 percent of the ocean but support nearly a quarter of all marine species, making them critical ecosystems under pressure from warming waters, overfishing, and coastal damage. Understanding where biodiversity concentrates is key to protecting them.

Sound and sight fusion

The CUREE system combines multiple sensing modes, including visual fish surveys, underwater sound mapping, and tracking of species behavior. This allows it to detect both visible marine life and hidden activity based on underwater soundscapes.

The robot can also navigate toward biological signals, such as snapping shrimp or fish calls, using sound-guided movement. It then confirms findings with close-range visual data, improving accuracy in complex reef environments.

In field trials conducted in the U.S. Virgin Islands between 2022 and 2024, the robot repeatedly identified a major biodiversity hotspot near a pillar coral structure at Joel’s Shoal. Fish density in this area was nearly 25 times higher than surrounding regions, supported by both acoustic and visual data.

“By combining these data streams, the robot can detect distant activity with sound and then verify that with close-up visual observations,” said researcher Seth McCammon.

The system also uses behavioral cues from key species to locate ecologically important areas. In one demonstration, it followed barracuda movement patterns to identify zones of high biological significance.

The robot successfully tracked underwater sound sources from distances of up to 80 meters and autonomously converged on reef hotspots from around 30 meters away. Researchers say this approach allows continuous mapping of reef ecosystems that are otherwise difficult to study.

Ultimately, the team envisions fleets of such robots operating globally to monitor reef health and guide conservation efforts as oceans continue to warm and ecosystems degrade.

“As coral reefs face unprecedented challenges, we need smarter, faster ways to understand where life persists and why so conservationists and resource managers can focus their attention where it’s needed most” said lead researcher Yogesh Girdhar.

Researchers say this approach could significantly improve long-term reef monitoring by reducing reliance on manual surveys and enabling continuous, high-resolution mapping of underwater ecosystems that are otherwise difficult, time-consuming, and expensive to study using conventional methods alone.

The study was published in Science Robotics.