Researchers at the Technical University of Munich (TUM) have developed a tool that mimics the intuitive motion patterns found in humans and animals to improve robotic movement. By tapping into natural oscillation and resonance, the tool enables robots to achieve energy-efficient, dynamic movement.
Humans and animals possess an innate ability to move with efficiency and ease, often unconsciously leveraging natural oscillation patterns to conserve energy. These patterns are linked to the body’s natural resonance, where movements are synchronized with a system’s inherent rhythms. Everyday actions like bouncing a ball or hopping on a trampoline are examples of this phenomenon, showcasing how our central nervous system (CNS) coordinates motion by tuning into natural frequencies.
In robotics, however, achieving such intuitive motion is a complex challenge, particularly in nonlinear systems. Unlike simple linear systems, nonlinear systems exhibit unpredictable dynamics but also feature periodic motion patterns known as nonlinear normal modes (NNMs). These modes provide a blueprint for leveraging oscillatory behavior even in chaotic systems.
Led by Albu-Schäffer, the TUM team adapted their computational tool to analyze and predict these natural oscillation patterns in robotic systems. This tool identifies energy-efficient movements by aligning a robot’s motion with its inherent oscillatory behavior.
BERT, a small quadruped robot designed by the German Aerospace Centre (DLR), became the testing ground for this concept. Researchers discovered six efficient movement patterns for BERT, including walking, trotting, and hopping. These movements utilized the robot’s natural resonance, allowing it to conserve energy while maintaining agility.
The team’s approach relies on delivering precise energy impulses through a computer-controlled regulator. This mechanism, akin to a parent pushing a child on a swing, ensures that movements align perfectly with the robot’s oscillation cycles. Unlike the intuitive human ability to align with rhythms, robots require mathematical precision to achieve the same effect.
The effectiveness of this strategy was showcased in a controlled race between three BERT robots. The robot programmed with the new method outpaced its counterparts, moving faster and more dynamically while using less energy. This result highlights the potential of rhythm-based motion control in robotics.