Enabling Data Driven Methods for State Modelling in Soft Robots

This thesis examines the challenges of creating accurate state models for soft robots. These robots, made of flexible materials, are useful for tasks like search and rescue or delicate surgery. However, their soft bodies make it difficult to predict their movements. This research explores new data-driven methods to better understand and model how these robots behave. This approach allows for more realistic simulations and could lead to improved control of soft robots for various real-world applications. This work offers significant practical value by enabling the development of more reliable and adaptable soft robots controllers.