How AI Can Assist In Coral Reef Restoration

EGYPT, RED SEA - APRIL 2017: Lyretail anthias

According to a new study, AI can be used to track coral restoration by comparing the sounds of healthy and unhealthy reefs.
The sounds of a reef come from a variety of sources, including the water washing over the reef and the animals living among them. An unhealthy reef would likely be less noisy, as fewer things would be nestled throughout. Thus, a team of researchers considered training a computer to distinguish the quality of a reef by its soundscape — something the human ear cannot easily detect.

Using the Mars Coral Reef Restoration Project in Indonesia, where reef spiders have been used to repopulate reefs destroyed by dynamite fishing, scientists deployed hydrophones for at one hour intervals across four kinds of sites: healthy, degraded, recently restored, and mature restored.

After feeding the recordings to a computer algorithm, the scientists determined that the computer could differentiate between the acoustic patterns of healthy and degraded reefs. Traditionally, reef health is assessed by sending divers down to visually survey the degradation of a reef and the animals living within. However, this method requires more energy and expenses than suspending a hydrophone over a reef.

"This is a really exciting development. Sound recorders and AI could be used around the world to monitor the health of reefs, and discover whether attempts to protect and restore them are working," said co-author Dr. Tim Lamont.

While this study demonstrated that the sounds between healthy and unhealthy reefs are easily distinguishable, identifying how the soundscape changes across these two endpoints could help describe the progression of coral reef restoration.