Investigation of smart soft robtic grasping with variable stiffness

Abstract

The past decade has witnessed huge progress in the field of soft robotics research. Compared with rigid-bodied robots, soft robots possess the characteristics of safety, adaptability, flexibility and compatibility benefitted from their inherent compliance. As a consequence, soft robotic technologies have been used for development of service robots, surgical robots, collaborative robots, etc. Nevertheless, further development of soft robots faces many challenges that must be addressed before their wide spread applications. The major challenges include: 1) inability to offer high stiffness/strength as rigid bodied robots do due to the highly deformable, less rigid materials that they are made of; 2) lack of integrated, low cost and robust position sensors to provide positional feedback; 3) lack of rapid and inexpensive fabrication methods. Aimed to address these challenges, variable stiffness soft robots, in particular, soft grippers or hands are explored in this thesis. Position feedback for soft robotic hands based on smart materials is also studied. Moreover, automatic fabrication processes (based on 3D printing) are developed for such grippers. Research in this thesis is expected to provide a new insight into the development of smart soft robots endowed with automatic fabrication. Variable stiffness could be realized mechanically in macroscopic structures from an engineering point of view. However, there is a trend to use materials whose intrinsic rigidity can be automatically varied in a reversible manner by electrical control signals in order to make a system simpler, more compact with superior performance compared with structure-based variable stiffness methods. Therefore, this thesis primarily focuses on tuning of intrinsic material rigidity for variable stiffness in soft robotics. Firstly, shape memory polymer (SMP) is investigated for variable stiffness applications owing to its large range of modulus change around glass transition temperature. By embedding SMP into the design, it is possible for soft robots to vary their stiffness via changing temperatures of SMP to above or below glass transition temperature. For rapid and automatic fabrication of SMP, polyurethane SMP is developed for 3D printing. A hyper-redundant robotic arm cascaded by a number of ball joints using SMP to modulate joint stiffness is 3D-printed as a case study. Combining SMP with soft pneumatic actuators, two versions of grippers are developed. The first version is an attempt in using SMP to modulate soft finger’s stiffness and thermal radiation is adopted to stimulate SMP. The second version uses embedded heaters to stimulate SMP. Conductive elastomers are designed into the soft finger with two intended functions: function as the embedded heater to modulate SMP and function as position feedback sensors for grasping control. Lastly, a soft hand with five fingers and a palm is designed and fabricated. Passive jamming is applied to vary the finger stiffness during grasping. Position feedback modules are also integrated into the design so that smart grasping strategies are made possible.