Soft Robotic Systems

Soft robots represent a groundbreaking shift in robotics, characterized by their flexible, deformable structures that enable them to interact safely and adaptively with unstructured environments and living organisms. Unlike traditional rigid robots, soft robots excel in tasks requiring delicate handling, adaptability, and versatility, making them ideal for applications ranging from medical devices to exploratory robotics.

In this project, we employ advanced manufacturing techniques, including 3D and 4D printing, to design and fabricate innovative soft robotic systems. These methods enable the creation of complex, multifunctional structures with tailored material properties and dynamic capabilities. For example, 4D printing allows us to integrate time-responsive changes into the robot’s behavior, such as shape morphing or adaptive stiffness.

Beyond the robots themselves, we focus on the development of customized soft sensors¹², which play a critical role in perceiving the robot’s environment and internal states. These sensors are designed to seamlessly integrate with the robot’s body, providing real-time feedback on parameters like pressure, deformation, and temperature.

The modeling and control of soft robots present unique challenges due to their highly non-linear and high-dimensional dynamics. To address this, we develop sophisticated algorithms capable of efficiently simulating and controlling these systems³. Leveraging reinforcement learning, we optimize their behavior by enabling the robots to learn from their environment and adapt to new tasks autonomously⁴. This learning-based approach enhances their performance in complex and dynamic settings, pushing the boundaries of what soft robots can achieve⁵.