Modelling of modular soft robots: From a single to multiple building blocks

Mohamed G.B. Atia,Abdelkhalick Mohammad,Andres Gameros,Dragos Axinte,Iain Wright
DOI: https://doi.org/10.1016/j.robot.2024.104622
IF: 3.7
2024-01-12
Robotics and Autonomous Systems
Abstract:Despite the advances in soft robots, their modelling is still one of the research challenges due to the complexity and non-linearity of their soft nature. A novel design of reconfigurable soft robots was recently introduced to bring more maturity to the field of soft robots. Here, a modelling study of the building blocks and the assembled soft robot is developed for a deeper understanding of the system and for predicting their behaviours. The model is validated using several sets of individual and multiple building blocks and the results show good tracking between the model and the experiments with reasonable errors. Towards adopting the model in robotic applications, a grasping setup of 2 assembled soft fingers is demonstrated where the model is used to predict the grasping force and compared to feedback sensing. The force prediction reveals the adaptability of the model and its robustness.
robotics,automation & control systems,computer science, artificial intelligence
What problem does this paper attempt to address?