Sharks from space 2025 NASA

.

the F.I.S.H. team

Predicting Shark Movements in Real Time

We built a framework that uses NASA satellite data and AI models to predict where sharks forage. Alongside, we designed a smart shark tag concept capable of detecting feeding events in real time.

An image showing the app we created for tracking sharks

Interactive Shark & Plankton Visualization Webb App

This interactive web application demonstrates real-time shark tracking visualization combined with oceanographic data. It is built with HTML, CSS, and JavaScript, and offers both 2D map and 3D globe views for exploring marine ecosystems.

Data Sources

  • Shark tracking data: Synthetic movement patterns for demonstration purposes.
  • Plankton concentration layer: Based on official NASA PACE satellite mission datasets showing chlorophyll concentrations in ocean waters.

Key Features

1. Dual Visualization Modes

  • Switch between a detailed 2D map view and an immersive 3D rotating globe
  • Both views show real-time shark movements with animated ping indicators

2. Interactive Shark Tracking

  • Click “Track Shark” on any individual to activate live tracking with red trail visualization
  • Tracked sharks display pulsing ping animations every 3 seconds
  • View complete journey histories with dashed yellow trails for selected sharks
  • Historical markers show first ping locations and feeding event detections

3. Sidebar Controls (toggle with menu button)

  • Search sharks by name or species
  • View live stats: current speed, heading, and last update time
  • Enable/disable the plankton concentration overlay
  • Sort and filter tracked individuals

Map Indicators

  • Blue dots: Untracked sharks
  • Red dots with ping animation: Actively tracked sharks
  • Yellow dots: Currently selected shark
  • Red markers: Detected feeding events
  • Color-coded overlay: Plankton density (low to high concentration)

How to Use

  1. Click the menu icon to open the shark list
  2. Select “Track Shark” to follow an individual with real-time updates
  3. Click any shark marker to view detailed journey history
  4. Toggle the plankton layer to see correlation between shark movement and food sources
  5. Switch between 2D/3D views using the button in the top-right corner

This mockup demonstrates how satellite oceanographic data could be integrated with biologging telemetry to study predator-prey relationships and migration patterns in marine environments.

Web Application Demo

Marine Species Tracker
Legend
Shark
Tracked Shark
Selected Shark
Journey Trail
Live Tracking
First Ping
Feeding Event
Chlorophyll Concentration
Low Medium High
Loading Earth...
Marine Species Controls
Plankton (21 locations)
Sharks (12 sharks)
Visible: 33
FPS: 60
Zoom: 1.0x
🖱️ Click & drag to rotate
🖱️ Scroll to zoom
🌊 Marine life in ocean areas
Tracker Family

Tracker Family

Framework that Infers Shark Hotspots

VERSION 1

Dataset Collection

Controlled environment data gathering for Edge AI model training

  • 9-DOF IMU: Accelerometer, gyroscope, and magnetometer for detailed motion tracking
  • Microphone: Captures bite sounds and feeding acoustics
  • Ultra-low-power accelerometer: Continuous monitoring at 1 Hz to detect activity
  • Hierarchical sensing: Low-power mode activates full sensors only when needed
  • Internal flash storage: Data retrieved after physical tag recovery

Primary Goal

Build a labeled dataset of feeding events in controlled environments to train and validate the behavioral detection model

VERSION 2

Real-Time Deployment

Edge AI-powered behavioral detection with satellite communication

  • Edge AI (TinyML): On-board feeding event detection in real-time
  • INS Positioning: Inertial Navigation System for 3D tracking between GPS fixes
  • Fastloc GPS: Quick surface positioning and INS drift correction
  • Pressure sensor: Accurate depth measurement underwater
  • Saltwater switch: Activates satellite link only when shark surfaces
  • NTN satellite link: Transmits summarized telemetry (location, feeding events, diagnostics)

Primary Goal

Detect feeding events on-board and transmit real-time behavioral data to continuously update predictive models

Tracker Evolution

V1 Tracker - Dataset Collection

V1 collects training data in controlled settings

Edge AI model is trained and optimized

V2 Tracker - Real-Time Deployment

V2 deploys trained model for real-time ocean monitoring

Challenge Summary

The challenge was to create a mathematical framework for identifying sharks and predicting their foraging habitats using NASA satellite data. It also involved designing a conceptual smart tag capable of detecting not only where sharks are, but also what they’re eating, and transmitting that information in real time to improve predictive models.

The goal was to uncover foraging hotspots and quantify how ocean conditions, phytoplankton communities, and predator movements are connected. As part of this, we explored new ways satellite measurements could help identify these crucial shark habitats.

F.I.S.H. Project — Framework that Infers Shark Hotspots

F.I.S.H. Project is a research initiative combining AI, mathematics, and satellite data to predict shark foraging hotspots and support marine conservation.

It consists of the Poisson Point Process (PPP) model enhanced with two inference strategies: a Gaussian Process for interpretability and uncertainty estimation, and a Transformer-based neural network for scalability and real-time learning.

Early results revealed that biogeochemical data alone can’t accurately predict shark behavior, emphasizing the need for behavioral insights. To fill this gap, we designed a conceptual smart shark tag equipped with Edge AI that can detect feeding events directly on the animal and transmit key data via satellite.

A web application complements the framework, visualizing real-time shark trajectories and feeding hotspots across the globe. Together, these innovations demonstrate how AI, edge computing, and satellite observations can work hand in hand to advance marine research.

The project is being developed in collaboration with the Department of Data Science at the University of Trieste, with plans for a scientific publication.

FAQs About Our Project

What is the main goal of this project?

The project aims to predict shark foraging habitats by combining satellite observations, advanced mathematical modeling, and real-time Edge AI tracking. This helps protect ocean ecosystems, inform conservation efforts, and support sustainable fishing policies.

Is the code publicly available?

Yes! The code and visual outputs are available on GitHub and in the project documentation.

What is a Poisson Point Process (PPP) and why did you use it?

A PPP is a mathematical model for random events scattered in space and time. It’s perfect for shark sightings recorded as individual points (day, latitude, longitude). We use an inhomogeneous PPP where shark density varies based on environmental conditions like temperature, salinity, and chlorophyll levels.

 

Why design a new shark tag when tracking devices already exist?

Existing tags provide location data, but we need behavioral data, specifically, when and where sharks feed. Our Edge AI-enabled tag detects feeding events on-board in real-time, which is critical for identifying true foraging hotspots and improving prediction accuracy.

 

What's the difference between V1 and V2 of the tag?
  • V1 (Dataset Collection): Designed for controlled environments to collect motion and acoustic data during known feeding events. It builds a labeled dataset to train the Edge AI model. Data is stored internally and downloaded after tag recovery.
  • V2 (Field Deployment): The production version with Edge AI, GPS, INS (Inertial Navigation System), and satellite communication. It detects feeding events on-board, tracks 3D position accurately, and transmits only summarized data via satellite when the shark surfaces.
How does the Edge AI detect feeding behavior?

The tag uses accelerometer data (ODBA – Overall Dynamic Body Acceleration) segmented into 2-second windows at 25 Hz. A lightweight convolutional neural network (CNN) trained via TinyML classifies shark behaviors like resting, swimming, and feeding based on movement patterns across three axes.

How does the tag save battery power?

It uses a hierarchical sensing approach:

  • A low-power accelerometer runs continuously at 1 Hz to detect activity
  • When activity is detected, the full IMU (9-DOF sensor) and microphone activate for high-frequency data collection
  • Data is transmitted only when the shark surfaces, using a saltwater switch to enable satellite communication

Resources

https://pace.oceansciences.org/data_images_more.htm?id=523

https://pace.oceansciences.org/data_images_more.htm?id=531

https://pace.oceansciences.org/data_images_more.htm?id=528

https://modis.gsfc.nasa.gov/data