6G Conference Berlin

We present our research platform MilleMani42 (MM42) as the enabler for post-humanoid robotics. Its versatile configurations allow application-tailored research and deployments for tasks requiring human-machine interaction as well as Physical AI.

This demo presents a 6G campus network integrated with Time-Sensitive Networking (TSN) for deterministic, time-critical communication. A prototype agentic controller uses an up-to-date Network Digital Twin to monitor performance, analyze network conditions, and automatically adapt the configuration.

The robotic hand can be teleoperated using a sensory exoskeleton. IMU sensors capture motion data, facilitating our precise, low-latency orientation tracking. The system bridges the gap between the motion complexity and seamless control and it demonstrates the capabilities of our real-time motion processing algorithm.

Experience a real 5G Standalone roaming setup featuring an O-RAN base station, commercial UEs, and interconnected mobile core networks. Attendees can interactively evaluate how security and network conditions impact connectivity and application performance.

An intent-based monitoring and control framework for next-generation networks using Digital Twins. Building on TwinRAN, the demo combines a cloud-based Digital Twin with an LLM-RAG pipeline to let operators monitor and control 5G networks through natural language. It demonstrates how high-level intents can be translated into intelligent network operations.

Lightweight generative modeling offers a way to capture the unique characteristics of the radio propagation environment in site-specific scenarios. Using these models, we demonstrate that the positional information of a mobile terminal is enough to infer significant features of the transmission channel’s transfer function, leading to highest performance.

Operators wear a head-mounted display showing a live stereoscopic feed from a quadruped robot, its camera tracking their head motion for natural 3D inspection of remote sites. But immersion depends on the network. Within 6G-life, we develop latency compensation and intention prediction algorithms to keep the view stable and the robot controllable.

Using the R2U2 (Realizable, Reconfigurable, Unobtrusive Unit) framework, the system continuously monitors packet-level telemetry transmitted over NB-NTN link and checks it against formal runtime specifications, flagging faults such as network congestion or link instability as they happen.

It is designed to support transformative experimental research across the ICT domain. This demonstrator functions as a sophisticated pilot environment, mirroring the architectural capabilities of the full-scale infrastructure. It is optimized for precise measurement and evaluation of 6G technologies, enabling fully reproducible low-latency measurements with nanosecond resolution.