Intelligent Wi-Fi
The earlier mobile device work generated some ideas on an “intelligent “adaptive interface - an interface that could be triggered to set off a series of preset actions by some environmental cue (such as time, location, or device usage). One of the potential methods to change the interface could be via wireless network; this has the advantages of being location driven, and usually tied to some aspect of lifestyle or activity.
One challenge of an adaptive approach based on wi-fi was designing around a location that could have different goals at different times or user intentions. Imagine if you developed an interface that shifted based on the wi-fi location at your local Starbucks; if you visited the Starbucks in the morning, you’d probably want to see your schedule for the day and your work contacts. In the evening, however, you might need to see your social contacts without your co-workers listed as well. The same locations, but vastly different user requirements - how do we design to address both?
This issue demanded a closer look at the potential limitations of task analysis; while analyzing tasks can optimize a design for a particular task, it cannot anticipate for scenarios in which the requirements of the task are violated (such as the Starbuck’s example above). Scenario-based design is similarly limited - the scenario must be imagined before the analysis can account for the action required. Imagine if one of the tasks failed for the wi-fi interface on the right - how would the user diagnose what had occurred?
Given this, an approach was required that could look at categories of information in manner that was independent of the actual event. The rationale for using such an analysis for interface design is straightforward enough; if we can identify the constraints on system behavior, we can design an interface that makes this behavior visible. This, in turn, should allow the user to effectively diagnose and address the problem.
But what analysis approach would provide this insight? The information below is a brief description of the project; for more information, see the slides or paper below.
Work Domain Analysis for Interface Design
Work domain analysis provides a systematic way to incorporate domain information into a design. By modeling categories of information (rather than instances of action), WDA provides an event-independent method of analyzing constant system purposes. By systematically analyzing the domains that the determine system constraints, we can display adaptive information on the interface that reduces the need for mental calculation or problem solving.
Using a WDA
WDA uses two major tools - the abstraction hierarchy and part-whole hierarchy (abstraction-decomposition space) for developing the information requirements for the device. After leveraging the abstraction-decomposition space to create information variables, we can use the abstraction hierarchy to identify the constraints that need to be represented in the interface.
These information constraints can be leveraged to create an adaptive interface; for example, the constraints of signal generation, signal strength, security, and associated network profile can be visualized on the interface. For an example, see the interface below (compared to the task based interface presented below it).
For more information and more examples of the final design, see the publications below.
Awards
2006 HFES Communications Technical Group - Student Paper Award