Validation and evaluation of a low-cost fabric-based tactile sensor in comparison with an sEMG sensor

Abstract

Tactile sensors play an increasingly crucial role in the monitoring of muscle behaviour during diverse activities. Regrettably, accessibility is restricted by the high cost of commercial Electromyography (EMG) systems. To address this limitation, an affordable fabric-based tactile sensor is introduced, ensuring precision and reliability while expanding potential applications. The development and evaluation of this innovative sensor are outlined in this paper, with a performance comparison to a prevalent sEMG sensor.

The conception, design, and development of our fabric-based tactile sensor are presented, and the evaluation phase involves dynamic exercises performed by volunteers, with EMG signals recorded from the Biceps muscle and Flexor Carpi Ulnaris muscle. Emphasis is placed on assessing the fabric-based tactile sensor's performance with the sEMG sensor.

Three performance indicators are employed to measure signal similarity. Consistently high performance is observed, with Spearman's correlation exceeding 0.642, an energy ratio of approximately 80 %, and Pearson correlation coefficients ranging from 45 % to 82 %. These findings signify a robust agreement between the fabric-based tactile sensor and the sEMG sensor, affirming the reliability and cost-effectiveness of our alternative.

While the advantages of the fabric-based tactile sensor are highlighted in our study, limitations of sensor placement, signal processing, and individual variations are acknowledged. Nevertheless, the research underscores the potential of this cost-effective sensor as a valuable tool in the analysis of muscle behaviour.