Fraunhofer Institute for Computer Graphics Research IGDAs part of the OTC-DaTA project, Fraunhofer IGD is developing scalable AI methods for live image analysis and for creating a digital twin for underwater missions. The goal is the efficient fusion and evaluation of heterogeneous sensor data for managing maritime missions and supporting new business models in marine technology.
The OTC-DaTA project combines state-of-the-art technologies in AI-based image and data analysis with innovative frameworks for the maritime industry. The focus lies on the development of a digital twin for the management of underwater missions and the practical implementation of machine-learning algorithms for analyzing complex data such as 3D point clouds and stereo videos. In the “Starfish” use case, AI models for estimating population density are developed and evaluated in field tests. The project results significantly contribute to the digitalization and efficiency of maritime applications and strengthen the innovative capacity of the Rostock location.
Project Description
1. Innovative AI Approaches for Underwater Research
Fraunhofer IGD is dedicated to developing scalable AI solutions for real-time image analysis and the integration of semantic visual-analytics methods into multisensor data evaluation. In the “Starfish” use case, an image-based AI approach for classifying and estimating the density of marine organisms is implemented. A custom-built, annotated dataset enables the transfer of algorithms to other populations such as sea fans or fish schools.
2. Digital Twins and Cloud Infrastructure
The project implements a digital twin of the “Artificial Reef Nienhagen,” enabling mission planning and evaluation in a VR environment. Data processing takes place via a container-based cloud infrastructure (OpenStack/Kubernetes), ensuring flexible and secure integration of sensor data and AI models. Interfaces to GAIA-X and the use of open standards promote interoperability and sustainable transfer of project results.
