Adapting smartwatch interfaces to hand gestures during movements: offset models and the C-shaped pattern of tapping

Abstract

Interacting with smartwatches is fairly common when users are moving. Although novel interaction gestures like flick of the wrist are implemented, basic touch gestures such as tapping and swiping still dominate. Using these gestures during a variety of movements could be challenging, and it is still not clear how the interface of smartwatches should tailor to users’ gestures during movements and how the usage of smartwatches influences the pattern of users’ movements. Therefore, this study investigates the interrelationship among users’ interaction gestures, movements, and gait features. An experiment was conducted among 47 participants, who used smartwatches through tapping, swiping, and wrist flicking to complete daily tasks in stand, strolling, normal walking, rushing, and jogging. They were tracked through built-in accelerometer and angle sensors. Four findings were derived from the experiment. First, rushing and jogging significantly decrease the effectiveness and efficiency of tapping. To reduce the tapping deviation, offset models were proposed and tested. Second, there is a C-shaped pattern on the round screen where tapping targets achieves higher accuracy than other areas. Third, the tapping performance could be improved by setting target sizes. Target sizes at 0.7 cm in stand, 1.1 cm in strolling, and 1.1 cm in walking achieve a high level of accuracy (95%), while target sizes at 1.5 cm in rushing and jogging achieve a middle level of accuracy (90%). Finally, tapping, swiping, and wrist flicking when users are moving significantly reduce their gait symmetry and step length. They do not imply significant influence on gait intensity, regularity, and overall stability.