Workload estimation of handwriting tasks using time series of pen pressure

Abstract

Pen comfort is an important factor in selection of pen for handwriting. In handwriting, a pen receives an impact when it contacts to paper placed on a desk. To hold the pen, fingers have to resist the impact. The greater impact, the grater mechanical workload on fingers associated with the grip strength to hold. Therefore, time series data of pen pressure is supposed to indicate the physical workload in handwriting task. As a preliminary step for the pen comfort evaluation, this study aims to evaluate the physical workload in handwriting tasks using time series of pen pressure. To judge whether pen tip contacts with paper, a probabilistic analysis method of time-series pen pressure was built first. Next, several indices of impacts such as average pen pressure and duration per stroke, and number of touchdowns per second were calculated. Finally, the product of the indices was defined as the normalized physical workload. For multiple four-character idioms with different stroke counts, pen pressure was recorded and the indices and the evaluation value were calculated. The results indicate that the normalized physical workload was larger in writing idioms with longer duration per stroke. In holding and moving a pen against the paper, it is considered that keeping pen pressure requires larger workload than resisting fluctuation of the pen pressure.

Keywords

• physical workload
• handwriting
• pen pressure
• binary bayes filter
• impulse
• four character idiom
• 6dof force and torque sensor