A Survey on AI-Driven Unified Data Platforms for Oceanographic, Fisheries, and Molecular Biodiversity Insights

Abstract

Marine life depends on shifting mixtures of physical, chemical, biological, and tiny-scale forces. To grasp how they connect, we need many kinds of info - like water patterns, fishing movements, plus species diversity clues. Lately, satellites, self-driving buoys, constant sensor grids, boat tracking tools, along with advanced gene reading methods have flooded us with oceans of fast-moving, varied data. Still, this information stays locked away in separate groups using mismatched structures, so pulling it together is tough. Smart computing can help glue these scattered pieces into one working system, opening doors to discoveries once out of reach.

This survey dives into AI-powered marine data systems, focusing on combining ocean readings, fishery records, and genetic biodiversity details. It looks at how these tools evolved, worldwide efforts, tech setups, data labeling rules, smart algorithms, plus hurdles in making systems work together. A five-part structure is suggested - covering gathering data, saving it, standardizing formats, analyzing insights, then delivering results. Real-world examples are included, along with missing pieces in current research, new hybrid AI methods, shared learning networks, linked knowledge maps, and upcoming ideas like virtual sea models and self-running DNA trackers. The goal? To help big national projects like MoES's Digital Ocean push forward by shaping India’s first complete marine data network.