Design of a low-cost sensor matrix for use in human-machine interactions on the basis of myographic information

Abstract

Myographic sensor matrices in the field of human-machine interfaces are often poorly developed and not pushing the limits in terms of a high spatial resolution. Many studies use sensor matrices as a tool to access myographic data for intention prediction algorithms regardless of the human anatomy and used sensor principles. The necessity for more sophisticated sensor matrices in the field of myographic human-machine interfaces is essential, and the community already called out for new sensor solutions. This work follows the neuromechanics of the human and designs customized sensor principles to acquire the occurring phenomena. Three low-cost sensor modalities Electromyography, Mechanomyography, and Force Myography) were developed in a miniaturized size and tested in a pre-evaluation study. All three sensors comprise the characteristic myographic information of its modality. Based on the pre-evaluated sensors, a sensor matrix with 32 exchangeable and high-density sensor modules was designed. The sensor matrix can be applied around the human limbs and takes the human anatomy into account. A data transmission protocol was customized for interfacing the sensor matrix to the periphery with reduced wiring. The designed sensor matrix offers high-density and multimodal myographic information for the field of human-machine interfaces. Especially the fields of prosthetics and telepresence can benefit from the higher spatial resolution of the sensor matrix.