Oakley, I., Lee, D., Islam, R., and Esteves, A. 2015. Beats: Tapping Gestures for Smart Watches. In Proceedings of the 33rd Annual ACM Conference on Human Factors in Computing Systems (CHI ’15). ACM, New York, NY, USA. [download]
Oakley, I., Lee, D., Islam, R., and Esteves, A. Method for Providing User Interface According to Beats Touch Based on Mobile Terminal, (10-2015-0021114), 11/02/2015, Filled.
Beats is a new type of multi-finger input that is specifically designed for the very small touch screens of smartwatches. It is based on what we term beating gestures, pairs of simultaneous or rapidly sequential touches (and optionally one or more releases) made by the index and middle finger of one hand. Essentially, instead of tapping a single finger to a screen, a beating gesture involves adjacent screen contact (and optionally release) with two fingers and in three closely controlled intervals: either simultaneously or with one event immediately preceding the other as part of a single coordinated movement. The timings and crude relative position (left or right) of this input serves to characterize each movement and the fast pace of the paired touches distinguishes it from regular single finger input or noise.
The motivations for exploring this specific design space include the fact that current smartwatch systems are large enough to register two adjacent touches and the simple sequences of movements required for beating gestures may be executable rapidly and eyes-free, avoiding fat-finger problems. Furthermore, we suggest the technique is sufficiently expressive to control a meaningful range of functions. Finally, as it is based on standard touch-screen input, but designed to not interfere with normal single finger use, we suggest it will be easy to integrate with existing on-screen interface styles such as single finger taps, swipes, dwell and gestures. Indeed, effective integration with established input techniques was a core design goal in our work.
In order to minimize on-screen clutter we designed systems that were effectively unmarked or invisible until evoked (in much the same way as swipe gestures are currently). To avoid fat-finger problems of users’ own bodies obscuring feedback, the designs relied on input (such as strokes) that could be made eyes-free or visuals that were purposefully displaced from a user’s on-screen finger. To explore the diversity of the technique we sought to design interactions based on a wide range of beating gestures. Furthermore, we purposefully integrated the beating gestures with traditional input styles such as menus, pie menus and sliders/scrollbars in order to explore how beating gestures could be combined with more established touch screen interactions, one of our key initial design goals. A final design concern was the consequences of inadvertent activation – we attempted to ensure that users would be able to cancel (or reverse) the outcomes of beating input rapidly and easily. In total, we created five prototype applications.
The prototypes were generally well received. Perhaps the clearest value came in repeated statements relating to the convenience, ease or speed with which the systems could be operated. Commenting on the menu shortcuts on the home screen P3 stated that usually there are lots of steps to access apps, but not here. Very fast. Similarly P7 liked that the zoom functions on the map were available right away and felt the alarm app was much quicker to operate than existing systems. P6 particularly appreciated the music player UI stating it would be useful even on a larger device such as a smartphone.
Users also tended to appreciate the interfaces that employed movements after the beating input. P1 stated that the strokes in the messages app pie menu were easy compared to having to touch some specific icon and controlling the alarm app was simple because it involved the finger still on the screen. Similarly, P9 appreciated the easy manipulation of the finger placement followed by vertical strokes during setting the alarm time. P10 indicated that the Dual-Tap followed by up/down movements to set volume was interesting and easy to use.
In sum, we believe this work showcases the potential of beating gestures for increasing the expressivity of touch input on small devices. Furthermore, it achieves this without requiring physical changes to the watch form factor, any fundamentally new sensing hardware (unlike, many prior approaches) or impeding or preventing existing interaction styles based on single taps, strokes, holds and pressure. We believe this demonstrates that beating gestures have the potential to be combined with these established techniques as standard elements of the touch screen interaction paradigm and look forward to exploring the novel interactions this enables.