Contents

Personal Digital Assistants


Personal Digital Assistants

A report Compiled by Michael Solomon

As an important group in the man-machine communication interface, Personal Digital Assistants today comprise a small but growing part of a relatively new method to accomplish tasks taken for granted on today's computers. The biggest attraction to PDA's is obviously portability. This feature also poses the greatest challenge to the designers of the human interface and inside hardware of these machines. The issues that arise while designing the human interface for small hand `computers' are similar in many ways to that for desktops. In this report, I will discuss a selection of the topics designers are faced with when designing applications that run on PDA's, with special emphasis on Apple's Newton PDA.

PDA's are not necessarily marketed to computer savvy consumers only. The design of applications would therefore include first time consumers as a potential user group, preventing designers from assuming that their users know some metaphors computer users are expected to. The major issues in human interaction with PDA's are screen real-estate, data entry, execution time for applications, maneuverability through screens, visibility of artifacts on the screen, communication with the rest of the world, user feedback on actions taken and actions pending, and handwriting recognition systems. Some of these parameters are discussed below. I did analysis on the subjects discussed by comparing the interface design problems encountered in class and in the project to my experience with PDA's.

Screen real estate

Here is one of the most important limitations of PDA's. In addition to the limited space on the screens, designers are usually constrained to a certain number of colors (if any) and gray scales. This makes the efficient design of icons, program interfaces, and other elements substantially more crucial. With the constrained space, the designers would also face the task of making all elements as friendly and as easy to use as possible for non-technical people, to make PDA's usable by most consumers if they are to replace their paper notebooks and organizers.

Data entry

Most PDA's today use pen-based computing schemes to let users perform a majority, if not all, of the featured functions. While using a pen fits the `paper-and-pencil' metaphor perfectly for most people, using it on a computer screen is a task not commonly done by many. To the first time user, introduction of a totally new method to interface with a computer could be challenging. My own experience with using the Newton for the first time showed me that to perform tasks taken for granted on PCs, for example highlighting and deleting a word in a word processing package, required tinkering with the interface for a while; a design parameter surely a of great consideration. With this comes a variety of similar tasks that designers have to adapt to fit the pen-computing model from what most computer users are accustomed to.

User feedback and graceful exit

To make PDA's look different than everyday computers, and therefore target the average consumer, designers look for ways to do things distinctly. Since PDA's generally take longer to perform tasks than their bigger and faster counterparts, greater attention is given to designing schemes to indicate when the system is doing time-intensive jobs. The Newton uses a flashing icon while other applications use a `Processing' or other similar text message to indicate a `busy' state.

For the same `slowness' factor that comes along with using PDA's, the ability to end an Application that takes a long time, or an application that is not responding, is perhaps more important here. Users perform short tasks and would probably be more irritated if the PDA refuses to, for example, take an entry in a schedule form.

Handwriting recognition

Since the theoretical maximum of 100% handwriting recognition is not yet attainable by today's machines, wise implementation of the available recognition technology is called for when designing PDA's. This could include requiring the user to learn a specific writing style, like US Robotics' Pilot PDA, or creating shortcuts that let the machine do most of the work that require input from the user. Word processing inherently calles for an optimum recognition package, but approaches such as a built-in keyboard could be effectively used to type in short commands.


Contents