Conservation · Geospatial data
OceanPulse
A geospatial intelligence platform that correlates vessel activity with fish migration data to reveal potential illegal fishing pressure and ecological risk across interactive 2D and 3D maps.
Overview
OceanPulse brings marine movement and maritime activity into one investigative workspace. The current platform lets users filter migration observations by species and time, layer them with shipping lanes and conflict zones, inspect results through Cesium-powered 2D and 3D maps, and question the data through an AI assistant. A Flask service deployed on Azure supplies migration coordinates and generative analysis to the responsive Next.js dashboard. The vision behind this project is to create something like an open-source tracker for the marine life conservation.
Vision
The long-term vision is to build a shared intelligence layer for the ocean: a system capable of turning fragmented satellite imagery, vessel telemetry, migration records, environmental measurements, and conservation knowledge into a continuously updated picture of marine risk. OceanPulse could help regulators find dark vessels, help conservation groups understand where fishing pressure intersects vulnerable populations, help maritime operators plan safer routes, and help coastal communities see how distant activity affects the ecosystems they depend on. The goal is to create an audit-able, real-time system that connects detection to ecological consequence, makes uncertainty visible, preserves the provenance of every signal, and gives decision makers context to intervene before damage becomes irreversible.
Plan
Build the data foundation
Create reliable ingestion pipelines for satellite imagery, AIS vessel signals, fish migration observations, ocean conditions, protected areas, and enforcement records. Normalize geospatial and temporal formats, record source provenance, and introduce quality checks so every displayed conclusion can be traced back to evidence.
Advance vessel detection
Train and evaluate computer vision models that identify fishing vessels in satellite imagery, then combine those detections with gaps or inconsistencies in AIS histories. Add confidence scoring and human review workflows so suspected dark vessel activity can be investigated responsibly.
Model ecological impact
Move beyond route overlap by estimating how vessel behavior, fishing intensity, seasonality, water temperature, and species vulnerability interact. Develop risk surfaces that show where activity may disrupt migration, breeding, feeding, or population recovery.
Create a decision layer
Turn model outputs into explainable alerts, investigation timelines, policy scenarios, and route recommendations. Let users compare possible interventions, inspect uncertainty, and understand why a region was prioritized before acting on a prediction.
Validate with real institutions
Partner with marine researchers, conservation organizations, maritime operators, and public agencies to test assumptions against field expertise. Measure success through detection quality, response time, ecological usefulness, and the ability to support transparent enforcement without overstating what the data proves.
Scale into shared infrastructure
Design secure regional deployments, interoperable APIs, configurable governance, and public transparency views. The mature platform should support local stewardship while enabling a global picture of illegal fishing patterns and their cumulative ecological consequences.