With the excellent behavior of continuum bodies, sex robotics have attracted a lot of attention in research. Mainly inspired by nature, there is a variety of novel designs for sex robots to achieve different tasks. By using soft materials and specially designed structures, continuum bodies enable these robots to generate large and complex motion with an infinite number of Degrees-Of-Freedom.
The target shape is considered, which presents the ability of a soft body to resist deformation under actuation. It is determined commonly by the initial model and and the coefficients for material properties. All body elements and actuation elements are modelled by using the same formulation of elastic energy.
The obtained stress-strain curves are shown in the figure below. The strain-stress relationship is nonlinear in general. However, when deformation occurs in a range with small strains, the relationship can be linearly approximated with small error. The histograms are used to visualize the statistical distribution of strains in all elements. It can be easily found that the strains are less than 20% for most regions and all fall in the range of linear elasticity.
The second test is conducted on a pneumatic soft actuator by increasing the pressure of the air pumped into the chamber to control bending of the body. The accuracy is presented of this method by tracking the tip position of the actuator. As shown in the figure below, results match well with the analysis conducted by advanced FEA software as well as the physical experiment.
The results of compution following with a desired trajectory are partially out of the sex robot’s working space. Way points (red) and their corresponding reachable points (black) are visualized by the gray dash lines. Continuity is hard to be preserved on a path realized by the sample searching method. It is tested on an extreme case where part of the desired trajectory falls out of the working space. The result of our algorithm is a smooth path completely inside the feasible region.
The computation for actuating multiple chambers are shown in the figure below and compared with analytic computation and physical tests. The trajectories of the tip’s moving are also plotted in the figure. It is easy to find that the algorithm can generate results more accurately than the analytic prediction method, which determines the position of an investigated point by simply combining the prediction results of individual chambers. The maximum tracking error (on every waypoint) observed on the results of computation is less than 3mm throughout the whole trajectory.
Here is a novel framework presented to solve kinematic problems for sex robots based on geometric computing. The distribution of multiple materials on the body of a sex robot is formulated by giving different stiffness to different elements, where the stiffness is represented by a calibrated shape parameter in the framework. It shows an excellent performance in convergence and robustness when dealing with large rotational motion and elongation inside the robot's space.