Information Technology Reference
In-Depth Information
Using force sensing as an input mechanism
(Scott et al., 2009) is another method that relies on
rather small movements. In this case, the sensing
of physical forces is applied by the user's hands
bending or twisting the case of a mobile device
that itself does not need to be more flexible than
a normal mobile phone to detect the meaningful
hand movements. Force applied to the casing is me-
chanically transmitted through it and parts attached
to it. Because of this, force sensors do not need to
be located at the external surface of the device,
and device cases “can be made with fewer holes
for physical switches, which can facilitate more
robust, more easily manufactured, and smaller
form factor devices” (p. 136). Although several
force gestures are possible, the paper concentrates
on the implementation of technology that allows
bending and twisting movements. The authors
implemented the two force-based shortcut interac-
tions on an Ultra Mobile PC (UMPC), where the
bend action is used to indicate 'page down', and
the twist action is used to change the foreground
application (as 'alt-tab' in Windows). Both pro-
vide suitable visual feedback to the respective
movement that mimic real-life interaction with a
window or document. Thus, for page-down, the
visual feedback looks like bending a topic and
flicking through the pages. The right-hand page
seems to flip up and over to the left hand side. Of
the 20 participants included in the user-based study,
10 were assigned to a 'bend', and 10 to a 'twist'
condition. The study employed a familiarization
phase, a training phase after which the device was
calibrated to each participant's preferred amount
of force applied to initiate the target action, and
a test phase. In the test phase, which was based
on Fitts' Law, the user had to move a force cursor
as quickly as possible to a special-coloured target
and then hold the force cursor inside the target
for two seconds. The number of targets in a block
varied randomly from two to eight, and blocks were
repeated in three cycles in which each number of
targets was attempted once, resulting in 210 target
acquisition attempts for each participant. Targets
varied in width and in distance from the starting
point. The 'bend' action appeared to be easier than
the 'twist' action in the sense that participants'
acquisition time was significantly shorter than in
the twist condition and it did not change between
the first and the third cycle whereas acquisition
time was significantly lower in the third- than in
the first cycle in the twist condition. In addition,
and contrary to the predictions of Fitts' Law,
acquisition time decreased in both conditions as
distance increased, with the closest targets being
most difficult to acquire. One major benefit of
this type of approach is obviously the saving of
screen real estate, giving more space to the actual
application.
The final study in this genre of gesture-based
interaction concerns a completely non-visual
interface that repurposes a touch screen as a
'talking' touch-sensitive surface, called 'Slide
Rule' (Kane, Bigham, & Wobbrock, 2008). Like
the studies above, it also requires somewhat finer
movements as well as integrating speech. Users
navigate through lists of items by scanning their
fingers down the device surface; gestures are used
to interact with on-screen objects. It uses a set of
four interactive gestures. The one-finger scan is
used to browse lists, for example, of phone num-
bers; users can select items on the screen with the
second-finger tap. They use a multi-directional
flick gesture to flip between pages of items or a
currently playing song, and an L-select gesture
to browse hierarchical information, for example
the hierarchy of artists and songs in a MP3 player.
As it provides speech output, this kind of technol-
ogy could be suitable for blind people as well as
for people with hand tremor or people who have
problems with fine-motor hand-eye coordination.
The paper reports a pilot study in which eight blind
users were interviewed to identify usability issues
with mobile devices and touch screens. Results
were used to inform the design of the Slide Rule
system. A new sample of 10 blind individuals took
part in the evaluation in which each performed
three tasks in each of five trials, namely (1) plac-
Search WWH ::
Custom Search