We operate data collection infrastructure and develop openly available software tools for maritime intelligence. We gather real-time tracking data from our AegeaNET sensor network and operational data from sea trials with our autonomous surface vessels, ROVs, and swarm fleet. Our analytical tools transform this data into actionable insights for traffic analysis, route optimization, and autonomous systems research. We contribute to a European data space for remanufacturing, and publish FAIR datasets.

Software

Maritime Data Analytics Toolbox

Transform raw AIS data into actionable maritime intelligence. Our openly available Python toolkit processes vessel trajectories to reveal traffic patterns, identify maritime routes, and support decision-making for port operations and maritime safety.

Key capabilities:

  • Smart filtering of erroneous AIS messages
  • Trip extraction
  • Traffic density heatmaps
  • Trajectory smoothing via Kalman filtering

View on GitHub

Raw AIS dataset

Dataset preparation

The toolbox provides a flexible pipeline to transform raw vessel trajectory data (AIS) into analysis ready datasets. Users can easily configure the filters and selection criteria, including noise filtering, downsampling and trip extraction to form datasets for the area and period of interest.

Traffic density heatmaps

Effective traffic visualizations

Easily generate effective heatmaps that precisely depict vessel traffic within the area of interest. The toolbox allows for the creation of density maps (TIFF images) based on the selected resolution and vessel types. Additionally, a series of density metrics are provided, representing the number of passing vessels, the different trips, the total time travelling in an area and others.

Multi-Modal Data Fusion Toolbox (MMDF)

A streaming engine that fuses heterogeneous sensor data in real time, developed at SmartMove within the CREXDATA project. The European Commission's Innovation Radar features MMDF, rates it Tech Ready and names the University of the Aegean as its key innovator (news).

Key capabilities:

  • High-throughput stream processing on Apache Kafka Streams
  • Spatial processing at scale with the H3 hierarchical index, joining live streams with static records such as shapefiles
  • Fusion logic designed visually, then run as lightweight microservices
  • Standalone at the edge or integrated into a central cloud hub

Core engine on GitHub RapidMiner extension on GitHub

Tutorial

A seven-minute walkthrough of the toolbox.

Data

AegeaNET: an advanced open sensor network

AegeaNET is our real-time maritime and aviation tracking network deployed across the Aegean Sea. The system integrates Automatic Identification System (AIS) receivers for vessel tracking with Automatic Dependent Surveillance-Broadcast (ADS-B) receivers for aircraft monitoring, creating a comprehensive situational awareness platform.

View live network map

AegeaNET coverage map of the Aegean

Network coverage

Our network spans critical maritime corridors throughout the Aegean Sea, covering major shipping lanes, ferry routes, and island connections. 10 AIS and 2 ADS-B receiving stations provide continuous monitoring of vessel and aircraft movements across the Aegean, with an archive going back to January 2025.

What sets AegeaNET apart is our commitment to open science and collaboration. We are actively sharing data with research institutions, maritime authorities, and academic partners worldwide. The raw tracking data feeds into our on-premises computing infrastructure, where it is processed and archived. Research partners can request historical extracts or live feeds; email .

The network's data streams are integrated with our distributed computing cluster, enabling real-time analytics, trajectory prediction, and maritime traffic pattern analysis. This infrastructure supports both our own research initiatives and collaborative projects with partner institutions.

What the data supports

  • Ground truth for validation: real traffic to test prediction models, collision avoidance and autonomous navigation against
  • Live and historical data: a real-time feed, and an archive since January 2025 for traffic analysis and seasonal studies
  • Ships and aircraft together: AIS and ADS-B tracks in one network, for situational awareness and sensor fusion work
  • A common benchmark: one dataset for comparing trajectory prediction, behaviour classification and anomaly detection models
  • Teaching and theses: live feeds and the archive for courses, student projects and theses, from diploma to PhD

Open datasets

All datasets follow FAIR principles: persistent DOIs, Creative Commons licences and documented formats on Zenodo.