TUM Researchers Harness Nature’s Rhythms to Revolutionize Robotic Efficiency

TUM Researchers Harness Nature’s Rhythms to Revolutionize Robotic Efficiency

(IN BRIEF) Researchers at the Technical University of Munich (TUM) have developed a tool that enables robots to replicate the energy-efficient movement patterns of humans and animals. By leveraging “intrinsic dynamics,” the natural oscillations that make human and animal motion economical, the tool allows robots to perform smoother and more efficient movements. This was demonstrated using BERT, a quadrupedal robot, which exhibited dynamic and faster motion when programmed with these natural patterns. The research, part of TUM’s work through the Munich Institute of Robotics and Machine Intelligence (MIRMI), aims to advance robotics while addressing sustainability and global challenges.

(PRESS RELEASE) MUNICH, 18-Nov-2024 — /EuropaWire/ — Researchers at the Technical University of Munich (TUM) have developed an innovative tool that enables robots to replicate the natural, energy-efficient movements of humans and animals. By harnessing the concept of “intrinsic dynamics”—the natural oscillatory patterns that underlie economical movement—the team has taken a significant step toward making robotic motion more fluid and energy-conscious.

The tool, created by Professor Alin Albu-Schäffer and his team, makes it possible to calculate the most energy-efficient movements for robotic systems. This groundbreaking work is demonstrated through BERT, a dog-like quadrupedal robot co-developed by the German Aerospace Centre (DLR) and TUM. Funded by the EU through an ERC Advanced Grant, the research explores efficient locomotion techniques and has identified six natural movement patterns for BERT, including gaits similar to walking, trotting, and hopping. These patterns align with the hypothesis that leveraging natural oscillations can lead to optimal energy use.

The team enhanced BERT’s capabilities by incorporating a computer-controlled regulator, which times precise impulses to mimic the intuitive energy inputs humans naturally provide during motion, such as a parent pushing a child on a swing. The result is a robot capable of more dynamic and faster movements compared to traditional programming approaches, as shown in a comparative race between three BERT models.

This research is part of TUM’s broader efforts through the Munich Institute of Robotics and Machine Intelligence (MIRMI). MIRMI focuses on integrating robotics, data science, and sustainable technologies to address global challenges, including health, through cutting-edge innovations in fields like machine learning, human-robot interaction, and soft robotics.

Publications:

The findings are documented in several academic works, including:

1. Annika Schmidt et al., Finding the Rhythm: Humans Exploit Nonlinear Intrinsic Dynamics of Compliant Systems in Period Interaction Tasks, PLOS Computational Biology, 2024. Read more here
2. A. Sachtler et al., Swing-Up of a Weakly Actuated Double Pendulum via Nonlinear Normal Modes, European Control Conference, 2024. Read more here
3. D. Calzolari et al., Embodying Quasi-Passive Modal Trotting and Pronking in a Sagittal Elastic Quadruped, IEEE Robotics and Automation Letters, 2023. Read more here

Video Demonstration:

A video showcasing BERT’s dynamic movements and its advantage over conventional methods is available here: Watch the video.

This press release was written by the author to highlight the innovative strides made by TUM researchers in translating the elegance of natural motion into robotic systems.

Media Contact:

Corporate Communications Center
Andreas Schmitz
presse@tum.de

Contacts to this article:

Prof. Alin Albu-Schäffer
Technical University of Munich
Chair of Sensor-Based Robotic Systems and Intelligent Assistance Systems
Alin.albu-schaeffer@tum.de

SOURCE: Technical University of Munich