Detecting and interpreting muscle activity with wearable force sensors

Abstract

In this paper we present a system for assessing muscle activity by using wearable force sensors placed on the muscle surface. Such sensors are very thin, power efficient and have also been demonstrated as pure textile devices, so that they can be easily integrated in such garments as elastic underwear or tight shorts/shirt. On the example upperleg muscle we show how good signal quality can be reliably acquired under realistic conditions. We then show how information about general user context can be derived from the muscle activity signal. We first look at the modes of locomotion problem which is a well studied, benchmark-like problem in the community. We then demonstrate the correlation between the signals from our system and user fatigue. We conclude with a discussion of other types of information that can be derived from the muscle activity bused on physiological considerations and example data form our experiments. © Springer-Verlag Berlin Heidelberg 2000.