University of Pennsylvania Develops Controllable Artificial Muscles to Enhance Precision in Soft Robotics

·by Henderson·AI
University of Pennsylvania Develops Controllable Artificial Muscles to Enhance Precision in Soft Robotics

Researchers at the University of Pennsylvania have developed a type of artificial muscle that enables more precise motion control for soft robots. This technology is inspired by the small muscles found in human hands and eyes, capable of producing controlled movements, maintaining intermediate positions, and coordinating with other artificial muscles. These muscles combine liquid crystal elastomers, graphene, and embedded copper coils to convert electrical signals into muscle-like movements. Demonstration examples include a robotic finger pressing individual piano keys and an artificial eye shifting its gaze through controlled motion, highlighting the technology's potential for fine motor skills in soft robotics.

Artificial muscles allow soft robots to perform precise movements by integrating flexible materials with built-in sensing and feedback control. The technology, named CoilLCE, is inspired by the small muscles responsible for fine motor control in human fingers and eyes. It combines liquid crystal elastomers (LCEs), graphene, and ultra-thin embedded copper coils to create artificial muscle fibers that can convert electrical signals into controlled mechanical movements. The research team developed these artificial muscles to address the limitations of traditional soft robot actuators. While conventional robotic motors can provide precise movements, they often require large components and complex mechanical configurations, making it difficult to replicate the flexibility and compactness of biological muscles.

Potential Applications of Artificial Muscle Technology in Robotics

Inspired by the small muscles in human fingers and eyes, researchers designed these systems. Liquid crystal elastomers offer a potential alternative as they change shape when heated. However, controlling their movement remains a challenge. When the material is heated, it contracts, while cooling allows it to return to its original shape. This makes it difficult to maintain a specific intermediate position without continuous monitoring and adjustments. "This is the first time feedback control has been demonstrated on LCE actuators, which is a necessary step for the future application of such artificial muscles in robotics," said Cynthia Sung, an assistant professor in the Department of Mechanical Engineering and Applied Mechanics at the University of Pennsylvania, in a statement.

To enhance performance, researchers combined liquid crystal elastomers with graphene. Graphene makes the final material softer and more resilient, capable of withstanding mechanical forces during interactions of the artificial muscles. The research team then embedded an ultra-thin copper coil within the LCE-graphene fibers. This coil serves two functions: it generates heat to activate the artificial muscles and acts as a sensor by measuring changes in electrical resistance as the material moves. These resistance measurements provide feedback about the muscle's state, allowing the control system to adjust heating power in real-time. By keeping the resistance near a predetermined target, the system can regulate contractions and maintain intermediate positions over extended periods.

This closed-loop approach eliminates the need for separate position sensors, reducing the number of components required to control the artificial muscles.

Future Directions and Challenges

Researchers also demonstrated feedback control of the LCE actuators, addressing significant challenges in making these materials suitable for practical robotic applications. The team showcased the technology using an artificial hand and a pair of robotic eyes. In the demonstration, the artificial fingers pressed individual piano keys, paused, and moved to subsequent notes. The robotic eyes utilized four pairs of opposing actuators to reproduce controlled gaze movements. This arrangement mimics biological muscles, where opposing muscle groups work together to produce movement. When one artificial muscle contracts, its corresponding muscle stretches, allowing the system to control motion without needing to drive every muscle simultaneously.

This configuration also reduces the number of control inputs required.

Researchers believe this technology has potential applications in humanoid robots, particularly in artificial faces that require many small actuators to work together to reproduce natural expressions. However, integrating a large number of artificial muscles into a coordinated system remains a challenge. The team plans to explore artificial intelligence and computer vision to manage more complex movements and interactions.

H
About the author
Henderson