Open6GHub+ – Flexible AI for 6G Networks

Open6GHub+ demonstrates how cutting-edge 6G research translates into practical applications. As an open transfer center, Open6GHub+ connects science, industry, and society, bridging the gap between basic research and real-world impact.

The demonstrator presents a flexible, distributed AI system for future 6G networks. Sensor data—particularly video streams—is analyzed in real time across multiple edge devices. Visitors can observe how the computational load dynamically adapts to device capabilities, data volumes, and network conditions.

It illustrates key 6G concepts, including the flexible shifting of AI processing between the edge and the cloud, scalable AI services, and efficient data processing at the network’s edge.

Interaction for Visitors

• Switching between different analysis pipelines
• Scaling AI processing in real time
• Distributed analysis across heterogeneous devices

Technology

The DiDet AI module uses modern object recognition. Batch inference processes multiple data streams in batches, saving computational power and energy. Edge compression reduces data directly at the source, lowering bandwidth requirements.
Rule-based middleware orchestrates, integrates devices with varying performance capabilities, and intelligently distributes data streams. The system also demonstrates MLOps functions, including performance monitoring, automated model updates, and stable AI operation across edge and cloud systems.

Open6GHub+ is part of the German 6G program and is funded by the Federal Ministry of Research, Technology, and Space (BMFTR). The goal is to strengthen Germany’s position as a leading market for the next generation of mobile communications and to advance trustworthy, resilient digital infrastructures in collaboration with industry, SMEs, start-ups, and public-sector actors.