Irainature Info

Researchers at the University of Queensland paired underwater drones with convolutional neural networks trained on millions of coral‑health images. The system can classify a coral colony’s health status (healthy, bleached, diseased) with 94 % accuracy, and then direct robotic “seeders” that attach coral larvae to the most promising spots. Early field trials show a in survival rates compared with manual transplantation.

Their philosophy is rooted in maximising the potential of even the smallest things, which led them to the concept of —turning waste materials into valuable new products.