sibility support, and many people presumed touchscreens could not be made usable for blind people. Slide Rule developed a set of gestures and the first finger-driven screen-reading techniques to enable blind people to access and control smartphone touchscreens.

We became aware from a personal communication in 2010 that Slide Rule inspired aspects of Apple’s VoiceOver screen reader for iOS. Indeed, Slide Rule’s finger-driven screen reading, swipe gestures, and second-finger tap can all be found in VoiceOver today.

Slide Rule exhibited the first three principles of ability-based design; it also exhibited the fourth and sixth principles, as its screen reader could adapt to the speed of users’ movements, tailoring its performance to theirs. The underlying principles demonstrated in Slide Rule have survived into today’s touchscreen systems.

Walking user interfaces. Today’s smartphones are portable but not truly mobile because they support interaction only poorly while moving; for example, walking divides attention, 24 reduces accuracy, 17 slows reading speed, 26 and impairs obstacle avoidance. 32 We conducted multiple projects to improve interaction while walking, focusing on people’s abilities while on the go.

In our early exploration of walking user interfaces, 15 we studied level-of-detail (LoD) adaptations, where the interface shown while a user was standing had high detail and the interface shown while a user was walking had low detail, with larger fonts and bigger targets. When a user moved from standing to walking and vice versa, the interface changed. We compared this adaptive interface to component static interfaces for both walking and standing, finding that walking increased task time for static interfaces by 18%, but with our adaptive interface, walking did not increase task time. We also found that the adaptive interface performed like its component static interfaces; that is, there was no penalty for the LoD adaptation.

In our subsequent project, called
WalkType, 12 we made mobile touch-
based keyboards almost 50% more
accurate and 12% faster while walk-
ing. Touch-based features like finger
needed. Although the Global Public
Inclusive Infrastructure (GPII), 34, 35 with
its cloud-based auto-personalization of
information and communication tech-
nologies, was formulated independent
of ability-based design, its objectives
are the same—enable interfaces to be
ideally configured to match each user’s
situated abilities.

The GPII is built on three technological pillars. 35 The second, “ auto-personalization,” is the one of interest here.d Its long-term goal is to ensure that any digital interface a person encounters instantly changes to a form that can be understood and used by that person. The GPII’s auto-personalization capability uses a person’s needs and preferences, which are stored in the cloud or on a token, to automatically configure the interface of each device for that individual. 34, 36 Its “one size fits one” approach is designed to help each person have the “best fit” interface possible. Since interface flexibility on current devices and software is limited, GPII auto-personalization uses both built-in features and assistive technologies (AT) (on the device and in the cloud)

 

d The two other pillars make it easy for people to determine what they need or prefer, ensuring solutions exist for everyone.

location, duration, and travel were combined with accelerometer features like signal amplitude and phase to train decision trees that reclassi-fied wayward key-presses. Walk Type effectively remedied a systematic inward rotation of the thumbs caused by whichever foot was moving forward as the user walked.

Performing input tasks is only one challenge while walking. Consuming output is another. In SwitchBack, 19 an attention-aware system for smartphones, a smartphone’s front-facing camera was used to track eye-gaze position on the screen to aid task resumption. For example, when a user was reading and looked away, SwitchBack remembered the last-read line of text; when the user’s gaze returned to the screen, that same line was highlighted to draw the user’s attention for easy task resumption.

These three walking user interfaces exhibited all seven principles of ability-based design to varying degrees.

Global Public Inclusive
Infrastructure

Ability-based design has been applied mostly at the level of individual systems and applications, but for greater impact, a new infrastructure that extends beyond the user’s own device is

Table 2. Seven principles of ability-based design, updated and revised from previous versions. 37, 38

Principle Description

Designer Stance (required)

Ability Designers focus on users’ abilities, not disabilities,
Accountability Designers respond to poor usability by changing
systems, not users, leaving users as they are.
Availability Designers use affordable and available software,
hardware, or other components acquirable through
accessible means.

Adaptive or Adaptable Interface (optional)

Adaptability Interfaces might be adaptive or adaptable to provide
the best possible match to users’ abilities.
Transparency Interfaces might give users awareness of adaptive
behaviors and what governs them and the means
to inspect, override, discard, revert, store, retrieve,
preview, alter, or test those behaviors.

Sensing and Modeling (optional)

Performance Systems might sense, monitor, measure, model,
display, predict, or otherwise utilize users’
performance to provide the best possible match
between systems and users’ abilities.
Context Systems might sense, monitor, measure, model,
display, predict, or otherwise utilize users’
situation, context, or environment to anticipate and
accommodate effects on users’ abilities.