Real-time sitting behavior tracking and analysis for rectification of sitting habits by strain sensor-based flexible data bands

Abstract

From human–machine interfaces to human–computer/robot interaction, the prevailing pattern has gradually become one of people remaining in a sitting posture while working. However, unhealthy sitting behavior seriously affects human health. This paper presents a novel tracking and analysis method of real-time sitting behavior. It is designed using a series of flexible wearable data bands, based on flexible stretchable sensors and pressure sensors (PSNRs). A flexible PSNR is fabricated using composites of carbon black, carbon nanotubes and silicon rubber, by a mixed solution method; it possesses a good property of pressure perception, for tracking sitting behavior. The sensors are accurately attached to human joints for accurate measurement of joint movement at the shoulders, elbows, wrists, knees, and waist. In this work, a new idea of real-time sitting behavior recognition is introduced and developed, based on a radial basis function neural network. Dynamic time warping is used to select candidates for dynamic sitting behavior and also to recognize postures by comparing the observed records with a series of pre-recorded reference data patterns. The solution deals simultaneously with real-time sitting behaviors as well as with multiple joints within the area of interest, to monitor the health level of the sitting behavior and to remind humans of sitting habits. The experimental results of the real-time sitting behavior tracking and analysis verify the effectiveness of the proposed methods. Additionally, undesirable sitting behaviors were gradually rectified and the sitting habit health levels of the participants were gradually increased.