SEACOP - Using Robotics & AI to save the oceans

Growing up in Dallas, I did not have a daily connection to the ocean. It was mostly during some weekend trips to Galveston or the Florida Coast. Each time I went on snorkeling trips, local tour operators often spoke about how fish populations and ecosystems are being impacted. From the shores – everything looks the same, the waves, the blue waters. But under the waters, things are changing slowly.

As I began researching, I found that global fish populations have declined by 90% since 1950, and today, almost a third of fisheries are considered overexploited. Before Overfishing became a prevalent issue, fishermen reported an astonishing abundance of catch in the water, but this has steadily decreased over the years.

Not only does overfishing lead to fewer fish, but it also causes bycatch, which is the accidental capture of non-target species in fishing gear. Every year, over 100 million animals die as bycatch, mostly sharks, turtles, and cetaceans. The Vaquita, a porpoise in the Gulf of Mexico, has been decimated by bycatch to such an extent that it is believed there are only 8-10 individuals left in the wild. It is one among many species devastated by the industry.

Overfishing also has substantial human consequences. Large-scale industrial fishermen often steal fish from smaller artisanal fishermen. IUU (Illegal, Unreported, and Unregulated) fishing accounts for a fifth of global catch and deprives small-scale fishers of their livelihoods. This exacerbated poverty and hunger especially in places like West Africa and Southeast Asia. IUU Fishing vessels are also known to engage in human trafficking and the use of slave labor. While efforts have been made to monitor Overfishing and IUU Fishing, many fisheries observers are bribed, harassed, or killed.

As I learned more about fisheries management, a pattern became clear: most monitoring systems focus on what can be seen from the surface—vessel location, reported catch, or human observers onboard. This gap between where monitoring is strongest and where fishing actually happens led me to develop SEACOP, an end-to-end fisheries monitoring system designed to operate below the surface. SEACOP combines a low-cost underwater submersible with computer vision–based object detection and motion tracking to identify both target catch and bycatch in real time. By analyzing underwater footage with deep learning models, the system generates detailed, automated reports of fishing activity that complement existing enforcement and conservation efforts.

To learn more about my project, join the webinar organized by my First Tech Challenge Robotics team, AREA 52, on Sunday, January 25^th, 5 pm -6 pm CT. Registration link on the event flyer.