Smolbot-VS: a soft modular robot with variable stiffness backbones

Abstract

This thesis introduces a novel mobile modular C-legged robot featuring a variable stiffness mechanism. The robot’s capability to adjust the level of compliance is achieved through a tendon-driven actuation system typically found in continuum robots. This system, in conjunction with 3D-printed soft backbones, forms the core of the robot’s modular design. A significant contribution of this work lies in developing and controlling the variable stiffness mechanism for its integration into SMoLBot. The robot design, in particular the backbone design was iteratively refined through repeated case studies, enhancing the robot’s ability to sense obstacles and optimize stiffness quantification using data from conductive backbones. The research also examined the advantages of modulating backbone stiffness, particularly in improving the robot’s performance in uneven terrain in rigid and soft configurations. The work in the thesis is concluded by detailing the implementation of a PI stiffness controller that leverages voltage feedback from the backbones to adjust the robot’s stiffness to intermediate levels. This controller allows the robot to autonomously operate variable stiffness mechanisms while demonstrating another potential use of soft sensors for mobile robots. The cyclic load tests established a correlation between voltage data from the backbones and stiffness, with force and displacement measurements confirming the relation across different stiffness levels.