Machine Intelligence and Robotics Lab

Towards Machine Intelligence

The aim of our work is to develop a new generation of autonomous, flexible and efficient robots by integrating model knowledge, exploiting system dynamics, taking individual hardware strengths into account and combining them with machine learning methods. The research focuses on design concepts for new, intelligent machines and robots, situation awareness and self-awareness as well as action and decision making to increase robotic and machine intelligence.

Design concepts for intelligent machines

• Biologically inspired designs, drive and control concepts for multi-legged walking robots
• Co-design of compliant actuators with force/torque-sensitive locomotion approaches and manipulation strategies
• Retrofitting and expanding systems and mobile robots with biologically inspired, multimodal sensors

Situational awareness and self-perception

• Identification and utilization of robot dynamics for highly dynamic, physical interactions
• Machine learning for self-aware autonomy based on multimodal 3D maps
• Event-based, neuromorphic perception for fast reactions

Regulatory approaches, actions and decisions for autonomy

• Physics-Informed ML approaches for advanced motion planning and execution
• Mobile manipulation strategies for walking robots
• Semantic mapping, motion planning and LLM-based decision making