Abstract:The oceans are a principal source of biodiversity, and with a global seafood market worth over $120B, they’re a crucial resource to almost half of the world’s population [1]. Costing society $23.5B annually, overfishing caused by illegal, unreported, and unregulated fishing (IUU fishing) contributes significantly to this depletion of fisheries. According to the World Wide Fund for Nature, IUU fishing “threatens marine ecosystems, puts food security and regional stability at risk, and is linked to major human rights violations and even organized crime.” In some locations, government-employed observers accompany boats to prevent IUU fishing [2]. However, even in wealthy countries, observers only monitor a minuscule percentage of fishing vessels. For example, in the expansive region of the Pacific Ocean from Indonesia to Hawaii, just 2% of fishing operations are monitored by observers. To combat the problem of IUU the experimenter developed an Edge Technology Based Artificial Intelligence System for marine protected areas (MPAs) using low-cost edge computing devices to track illegal fishing activity through AI-based image recognition services. The product is a solar-powered, inexpensive, edge computing and monitoring device mounted on buoys with a video camera and processor to analyze images using machine learning models. The model detects vessels, monitors their illegal activity in the oceans, thus reducing the overexploitation of fishing. The edge device does processing locally and sends relevant data to the database, reducing the need for processing vast amounts of images & videos centrally. A stealth Autonomous Aerial Vehicle (drone) with a pre-programmed flight path collects the data from buoys and reports predictions to ground stations providing 24x7 surveillance capabilities.The product has a broad range of potential applications to detect overfishing, piracy, smuggling, and instances of ocean pollution, including oil spills. It can also be deployed for marine surveillance, primarily supporting the national defense. The immediate application for this product is the continuous surveillance and protection of targeted MPAs by alerting illegal fishing activities to governments and NGOs in real-time.

Predicting Illegal Fishing on the Patagonia Shelf from Oceanographic Seascapes

Anticipating Illegal Maritime Activities from Anomalous Multiscale Fleet Behaviors

Fishery catch records support machine learning-based prediction of illegal fishing off US West Coast

Marine Piracy Prediction and Prevention: Policy Implications

Zinc induced apoptosis in HEP‐2 cancer cells: The role of oxidative stress and mitochondria

Identification of suspicious behavior through anomalies in the tracking data of fishing vessels

An Open-Source Platform For Underwater Image And Video Analytics

Extracting Mediterranean Hidden Fishing Hotspots Through Big Data Mining

Identification of suspicious behaviour through anomalies in the tracking data of fishing vessels

Fine spatio-temporal prediction of fishing time using big data

Responding to Illegal Activities Along the Canadian Coastlines Using Reinforcement Learning

Detecting communities at high-risk of IUU fishing: networks of shadow encounters in Area 81 of the Western Central Pacific

Enhanced Fishing Monitoring in the Central-Eastern North Pacific Using Deep Learning with Nightly Remote Sensing

FAD-SAR: A Novel Fishing Activity Detection System via Synthetic Aperture Radar Images Based on Deep Learning Method

Spatio-temporal trajectory data modeling for fishing gear classification

A Hybrid Framework for Maritime Surveillance: Detecting Illegal Activities through Vessel Behaviors and Expert Rules Fusion

Towards a responsible machine learning approach to identify forced labor in fisheries

Predicting Fishing Footprint of Trawlers From Environmental and Fleet Data: An Application of Artificial Neural Networks

Website visits can predict angler presence using machine learning

Edge Technology Based Artificial Intelligence System for Ocean Patrol and Surveillance

Can machine learning predict citizen-reported angler behavior?