Cybernetic suit for interaction with virtual reality

Abstract

The study in the field of the structure of the processing architecture of a cybernetic suit for working with virtual reality was conducted. Nowadays it is relevant to use such a systems to implement various scenarios of rehabilitation for people, as well as to conduct possible training sessions for operators of mechanized devices and so on. The goal is to achieve the highest possible accuracy in repetition of the operator's movements, i.e. motion capture, at the lowest financial cost. The initial variant of the suit was created on the basis of Arduino and Unity 3D for visualization. 3D human model based on quaternion is built in Unity 3D which repeats all the real person twists and turns. But the first architecture had problems with performance and accuracy. A series of experiments was carried out to improve the effectiveness of the cybernetic suit prototypes. Thanks to the introduction of the proposed new type of architecture, namely the “controller-sensor”, it is possible to completely get rid of the inaccuracy of the received data from the sensors. Also it allows removing the unauthorized change in the position of the body parts of the 3D model and increasing the overall resistance and noise immunity of the product. Software based on the algorithm for processing data from gyroscope and accelerometer sensors is presented, and the possibility of upgrading the design of the hardware part of the suit, aimed at optimizing the output characteristics, is described. To achieve the goal the most stable and noise-resistant transmission of primary data from cybernetic suit to a 3D model with the minimum possible number of errors in the data transmission buses was obtained. As a result, the modernization of the cybernetic suit for interaction with virtual reality has increased the accuracy of data from 66 % to 94.5 % and reduced the amount of interference from 40.59 % to 5.5 %.