A form-finding method for deployable tensegrity arms and inverse kinematics

Manipulator arms in robots can be bulky and difficult to transport. Tensegrity mechanisms, which can be compact, deployed, and shaped to adjustable lengths, offer a promising alternative for robotic manipulators. This work develops a methodology for class 2 tensegrity mechanisms formed by cylindrical modules, using nonlinear programming to design a deployable tower capable of complex shape transformations, such as bowing. The study starts by deploying the structure from a compact shape into a tower, thereby enhancing its transportability and impact resistance. Next, a form-finding procedure assigns a bowing movement to the tower by pulling specific cables, using a kinematical method and nonlinear programming to achieve a stable configuration. Finally, the workspace of the mechanism is approximated through surface fitting, and three inverse kinematics functions are defined using artificial neural networks, sequential quadratic programming and a genetic algorithm. The example presented uses six quadruplex modules, but the floor and ceiling functions applied make it valid for any cylindrical tensegrity stacking.