Automatica 2025 in Munich showcased the latest advances in robotics and automation technology. DLR made a strong impression with our expansive booth spanning two areas in Hall B4 (Stand 321/324), where we demonstrated cutting-edge robotics and AI integration.
DLR’s Robotics Showcase Link to heading
Our Institute of Robotics and Mechatronics presented five groundbreaking systems:
TORO - Our bipedal humanoid robot demonstrates advanced locomotion, climbing stairs and navigating obstacles with enhanced perception capabilities.
AIMM (Autonomous Industrial Mobile Manipulator) - Showcases autonomous handling and transportation in industrial environments.
EDAN (EMG-controlled daily assistant) - An assistive robot combining a wheelchair with a lightweight robotic arm for people with severe motor disabilities.
SARA - Features my latest research combining Foundation Models (LLMs) with machine learning approaches (KMP) for adaptive robot motion.
HUG - A haptic interface enabling intuitive robot control through force feedback.
Interview Spotlight Link to heading
During the event, I had the opportunity to discuss my latest research combining Foundation Models (LLMs) with machine learning approaches (KMP) for adaptive robot motion on SARA. This innovative framework enables online adaptation of robot motions while maintaining safety, as the LLM doesn’t have direct access to the robot control—unlike other neural network approaches.
At the fair, I demonstrated a learning-from-demonstration skill where SARA picks up a bearing ring and places it inside a measurement station. After teaching the robot through a few demonstrations, we can adapt the motion through natural language commands—making it faster or slower in specific areas, adjusting height, or avoiding certain objects. This represents a significant advancement in safe, intuitive robot programming.