Creando la mayor base de datos de plástico flotante en el océano utilizando cámaras inteligentes

The Ocean Cleanup’s Automated Debris Imaging System (ADIS) has completed its first five years of operation, marking a milestone in the effort to catalog and track floating ocean plastic. Published in Environmental Research Communications, the findings reveal that ADIS has scanned over 14,500 square kilometers of the global ocean surface and identified more than 20,000 large floating objects. The system, deployed aboard vessels traversing major ocean basins, has generated more than 27 million photographs, equivalent to over 165 terabytes of data, since its launch in 2019. ADIS is designed to address the limitations of traditional methods for monitoring ocean plastic. Classical techniques such as surface trawls and visual observation struggle to capture the vast distribution of floating marine debris, particularly large macroplastics measuring more than 50 centimeters. These items, though relatively rare, contribute significantly to the overall mass of oceanic plastic. To overcome these challenges, The Ocean Cleanup developed ADIS, leveraging smart cameras and artificial intelligence to systematically scan and analyze ocean surfaces. Initially, the first-generation ADIS setup involved mounting a GoPro camera on the side of a vessel and operating it in time-lapse mode. Images were stored on hard drives and later transferred to Rotterdam for processing. Over time, the system evolved to incorporate advanced image recognition algorithms, allowing for automated identification of potential plastic objects. Each detected item was subsequently verified by human analysts, ensuring accuracy in the final count. From 2019 to 2024, ADIS operated across multiple ocean regions, including the North and South Pacific, North and South Atlantic, and the Indian Ocean. The data collected during this period has been compiled into a global dataset and interactive map, providing insights into the spatial distribution of ocean plastic. This open-access resource enables researchers to study patterns and concentrations of marine debris, supporting initiatives aimed at cleaning up the world’s largest garbage patches. A key challenge in the early stages of ADIS deployment was its accuracy in detecting floating debris. During a ground-truthing expedition to the Great Pacific Garbage Patch in 2022, comparisons between ADIS data and physical trawl samples revealed a discrepancy, with ADIS undercounting debris by approximately 5 to 6 times. This discrepancy stemmed largely from color and material biases affecting the system’s ability to recognize certain types of plastic. Despite this limitation, the consistent pattern of undercounting allowed researchers to develop a correction factor, improving the reliability of future assessments. Building on these insights, The Ocean Cleanup introduced ADIS2.0 in 2024. This upgraded version features autonomous capabilities, enabling real-time recognition of floating debris and automatic capture of high-resolution images with GPS coordinates. Unlike its predecessor, ADIS2.0 requires minimal human intervention, making it ideal for extended deployments. The system also streamlines data processing and transmission, reducing reliance on manual oversight. Currently, ADIS2.0 continues to collect data globally, surpassing the volume gathered by ADIS1. The ongoing accumulation of information is being integrated into an online data collection map, offering a dynamic visualization of ocean plastic distribution. With continued expansion, ADIS2.0 aims to generate periodic maps that reflect changes in plastic concentration over time, facilitating long-term trend analysis and informed decision-making. As The Ocean Cleanup refines its approach, the ADIS project stands as a critical tool in the fight against ocean pollution. By combining cutting-edge technology with rigorous validation processes, the initiative contributes to a growing body of knowledge essential for addressing one of the most pressing environmental challenges of our time.