Dimensionality reduction of cellular actuator arrays using the concept of synergy for driving a robotic hand.

This paper presents a method of reducing dimensionality of cellular actuator arrays. The cellular actuator arrays are used to drive a multi-axis system, i.e. robotic hand. The actuator arrays use Segmented Binary Control (SBC) which simplifies the control of nonlinear artificial muscle actuators. Although the SBC simplifies the control, the number of cells required to create motions is increased dramatically, thereby increasing the dimension of the actuator arrays. Therefore, reducing the number of controls, or dimension is needed. The coordinated motion of the robotic hand enables the coupling of actuators. In biological systems, synergies, a strategy of grouping output variables to simplify the control of a large number of muscles and joints is used to explain the coordinated motion created by the muscles. Similarly we can apply this concept to group the cells of the actuator array to be turned on or off together. Each group of cells, called segments can be designed to perform a certain set of desired postures. Data from fifteen different tasks is used in the design. The gathered joint angle data is transformed into actuator displacement data and used to generate a segmentation design of the actuator. For segmentation design, feature extraction method, similar to non-negative matrix factorization with binary filter is proposed. The segmentation design reduces 96 separately controlled segments to 6, while maintaining the ability to accomplish all desired postures. A prototype robotic hand with five fingers, designed and fabricated using the FDM process, is driven with this actuator system.