Mathmod 2018 Extended Abstracts

Parametric Reduction for Simplified Posture Classification Model

MATHMOD 2018 Extended Abstract Volume​, ARGESIM Report 55 (ISBN 978-3-901608-91-9), p 29-30, DOI: 10.11128/arep.55.a55225


A wide range of musculoskeletal disorders are caused by prolonged static sitting postures. As current posture recognition methods are mainly based on visual systems, their implementation at workplaces is nearly impossible. In this work we present a biomechanical approach for elimination of image based variables and development of a mathematical model requiring minimum possible inputs to classify the type of posture subject assumes. The model requires only the provision of only two subject specific parameters - height and weight and uses a single force input (as a time series) - the ground reaction force (GRF) and its point of application known as the centre of pressure (CoP), it automatically detects four classes of sitting postures defined by purely using the sagittal plane projection. The next step was to develop a database linking segment positions and mutual angles with the CoP position. Due to its simplicity the model was implemented for airplane seat and for in office environment where prolonged seating and the observed activities for the developed posture classification normally occur.