The ICT Bridge between Berlin and Istanbul for Research and Innovation

Aims at a holistic approach to support the streamlined development and deployment of data-driven applications towards realizing distributed autonomous systems, focusing on making device data available for AI/ML applications.
In this context, we have been developing an “AI middleware” and associated tools, which support both the extraction and semantic abstraction of device data as well as the development, deployment and monitoring of AI solutions on the ingested data.

Provides the necessary run-time and testing environment for AI models, libraries of trained AI models and data sets, as well as the supporting tools for general planning, perception and prediction functions.
A particular research focus of GT-ARC is on the design and development of the algorithms and tools that support more automated ML model selection and hyper-parameter optimization. We are also increasingly more involved in enhancing the explainability of AI models in our research, improving their applicability in critical domains such as automated mobility and health.

Focus is on the design and development of novel algorithms and tools to support network vulnerability monitoring and analysis as well as AI-based attack prediction, detection and response for connected systems.
In this direction, we have been developing both a real test environment and a simulation framework for monitoring and visualizing the network vulnerability state, attack modeling, and possible countermeasures. These are then utilized for game-theoretical modeling of attacker-defender interactions and the identification of countermeasure tasks, algorithms for automated response selection, and metrics for countermeasure effectiveness.

Diversity and complexity of problems require solutions with inherent intelligence such as monitoring its environment, processing the information, collaborating with others and autonomously making decisions.
GT ARC focuses on basic research activities for developing algorithms and approaches with self-* properties such as self-organization, self-healing and self-stabilization. The focal points of autonomous solutions are reliability, availability and adaptability.