Personal assault on the senses

Pocket-sized computers which can fax messages, read email and be used as a diary are now available but how relevant or welcome are they? Peter Thomas examines the issues.

A slick-looking executive, anonymous behind Ray-Bans, stares directly into the camera. He is seen slipping out a small computer from under the lapel of his tailored suit. "The road warrior's weapon of choice" says the caption in the Hewlett-Packard palmtop computer advert.

In the 1980s, when computers were appearing on everyone's desks, the move was towards wiring them up into local-area networks. Now there is a move away from the wired computer to the individual, mobile, wireless personal computer. It can be slipped into a pocket or briefcase, can be used to send faxes or read email from a car, or to take impromptu notes.

The most familiar example of this new form of very personal computer is the Apple Newton - the first "Personal Digital Assistant". The Newton, which enables users to make handwritten notes, use their diary and send faxes, also has an "intelligent assistant" which "works the way you do".

Originally launched with huge publicity, the Newton received what can only be described as disastrous reviews. This was mainly because of the mistaken assumption that simply putting together the capacity to make handwritten notes on a small screen, to beam an infra-red signal to another device, or to record diary entries, would go any way towards addressing the complexities of dealing with personal information.

As with all forms of technology, its introduction changes the environment into which it is introduced, creating new opinions, tensions and problems. One of the issues that devices such as the Newton have served to raise is the extent to which users actually want computers to manage personal information such as diaries or telephone books.

What devices such as the Newton have done is to make a consumer reality out of what was before only the dream of Alan Kay, a scientist at Xerox Palo Alto Research Center, who more than 20 years ago suggested a personal "Dynabook" which combined the properties of pencil and paper with computer power. As a result new PDA devices are appearing such as the Simon cellular telephone developed by BellSouth Cellular in the United States, which allows users to make calls, write notes and send faxes, and Bell Northern Research's Orbitel Personal Communicator - a small wireless device which will be able to accept telephone messages or faxes.

Personal Digital Assistants are only one aspect of the move to more personal technology. Xerox PARC, responsible for some of the most innovative computer applications, has been working on the development of "ubiquitous", everyday, computers. These small, portable devices allow users to move freely around their workplace connected only by infra-red and cellular wireless links.

One such device, the Active Badge, tracks the user's movements around buildings and can relay short messages or forward telephone calls to the nearest extension. Devices such as the Olivetti Active Badge are, suggest Xerox, better seen as "memory prostheses", designed to help users manage important information, than the future tools of an Orwellian nightmare. And, making the computer even more personal, is work carried out at MIT's Media Laboratory in the US. Here the Personal Information Architecture Group is recreating a living room inside the laboratory which can be used to develop devices embedded in everyday objects, and all linked to the "BodyNet" - wearable computer devices which allow the users to become one with their information environment.

All of these devices - from the PDA to the wearable computer connected to the BodyNet, and of course to the Internet - are opening up new phenomena for study and, correspondingly, new technology research and development problems. For example, one issue presaged in the Active Badge is sensitivity to the ownership of personal information (in this case physical location); an issue raised by the PDA is the way in which personal devices, once disconnected and mobile, can be reconnected to networks of other users.

It has been caustically remarked that buying a device such as a PDA is like buying a telephone 100 years ago - there is no one to talk to. A further issue is the way in which this personal technology will change the ways in which users work with information in their offices. Since the paperless office is unrealisable, how will paper and personal devices interact in the office of the future?

At the Centre for Personal Information Management at the University of the West of England we are looking at many of these issues. We are particularly interested in how people manage their information, and in how devices such as Active Badges, PDAs and computer-telephone integration applications can be designed to take into account the variety of working practices and information needs and how new forms of technology will change them.

One of our studies concerns the ways in which office-based workers - in particular managers and their assistants - work together to manage information. Through a video-based study of managers and assistants in a number of organisations we have found that there are many "information management" technologies used in the course of an ordinary working day: in addition to information from fax, telephone, voicemail, email and cellular telephone, it is not unusual for senior managers to run a variety of diaries - a personal electronic diary, a personal paper-based diary, a shared paper or electronic diary, and a separate diary for the manager's assistant. All of these are kept in step with information from the various information management technologies and information is dealt with, filed for future reference or passed on for action. How this is accomplished - the skills required to coordinate the information and the technology - are at present complex and little understood.

One of the major issues we have looked at is "time-management". One of the obvious difficulties is that of maintaining any coherent and up-to-date version of a diary split among several media - paper-based and electronic, or mobile and desk-based PC. There is a great deal of "coordination work" around technology, which from one perspective is occasioned by the technology (manager and assistant jointly spend a great deal of time comparing diaries and transferring notes and meeting dates). Our studies suggest that this coordination work is not simply an overhead in "getting the work done" but that it is the essential work of "keeping in touch with things" - regularly talking and reshaping appointments and meetings to arrive at a smooth working relationship and the smooth management of time.

In our view, the design of technology to support these relationships is not a matter of automating the tasks of a personal assistant (as the PDA does), but "informating" them - providing technology to support and enhance work. In addition to using our studies to uncover the complexities of information management, we are also using them to drive the design of new kinds of personal technology. One of these is a prototype "Personal Office Support System" - a device which sits on the desk of manager and assistant and is linked across a network or telephone line. It is centred on one of the most important information management tools - the diary - but allows the manager and assistant to shift information around in ways which support their working relationship. For example, telephone calls are diverted depending on the manager's booked appointments, as are faxes and email messages; the device maintains a set of "personal preference numbers" which allows some callers to bypass an assistant and call direct to the manager; the diary does not require detailed notes to be made of a meeting, but the manager can - using pen or voice input - simply block out slots of time as busy or free.

By way of experiment, we have also been looking at other kinds of personal information management devices that might be possible. One of the themes of the "ubiquitous computing" work carried out at Xerox PARC is the effect of scattering computational power around the user's environment. We have developed a prototype system called eDoor which allows users to indicate their availability not only to specific enquirers, but also to casual visitors. eDoor is a small flatpanel display which sits on the outside of an office door and indicates whether the door's owner is out, busy or available. On finding the door owner is out, casual callers can interrogate the door to find out where the owner is and when they might be back; they can also leave a voice and video message, or book a meeting through the eDoor's connection to a computer-based diary. Although not intended as a serious application, eDoor has been useful in allowing us to explore the links between computers that are everywhere, the information needs of users, and the demands of building robust and useful software and hardware devices to help people manage information.

One of our continuing research themes is that most ubiquitous of information management tools, the "personal information manager". PIMs are software applications which provide a diary, possibly attached to a personal database. One interesting observation here is the way in which these electronic diaries are simply translations to a computer screen of the physical look of a paper-based diary: time-slots, marked sequentially in hours and minutes into which users can enter appointments. We have found that users of such systems experience many problems with these applications however "well-designed" they may be. Expecting the electronic diary to make them more efficient, effective managers of their time, users instead find that the situation is often more chaotic than before.

As with our manager-assistant studies, we have found that users often maintain both a paper-based diary and an electronic one; users are typically extremely nervous about committing all their personal information (such as telephone numbers) to an electronic medium. Users also find that electronic diaries, although they have the same look, do not have the same feel as a paper-based diaries. For example, in many electronic diaries one can easily delete an entry - however the entry is then gone forever; in contrast using a paper-based diary a crossing out can often turn out to be valuable information and can be deciphered later. This of course only points to the difference between paper as a medium and a digital medium. But the issue is more than this, since paper is resilient and "trustable", and its use, along with diaries, is embedded in our culture. The challenge for the design of technology is therefore not to try and make one look like the other, but to design personal information management technology which combines the best of both media.

As is increasingly recognised, the quality of the user interface to computer systems largely determines their effectiveness, and so their success as commercial products. For these new forms of computing device, not only their user interface but their design as everyday personal appliances is all-important. As our studies suggest, the challenge for the design of these personal computers is that of harnessing novel technologies (such as handwriting recognition, infra-red and radio wireless links, voice recognition and small, flat displays) and applying them to the user's needs. The issues are not technological, psychological, or sociological, but a synthesis of these; and the adequacy of solutions cannot be judged by simple benchmarks of technological performance. Designing the "road warrior's weapon of choice" is an difficult, and personal, matter.

Peter Thomas is professor of information management and director of the Centre for Personal Information Management at UWE, Bristol. Contact: Peter.Thomas@csm.uwe. ac.uk / CompuServe 100255,2506 / http://gate.uwe.ac.uk:8000/csm/research/PIM/index.html More information appears in The IEE Symposium on Developments in Personal Systems: IEE, Savoy House, London, Friday 23 June 1995 (IEE:071 344 5419). Mobile Communication and Collaborative Technology, (edited by Peter Thomas) to be published by Alfred Waller/Unicom Seminars Ltd in 1995. The Social and Interactional Dimensions of Human-Computer Interfaces (edited by Peter Thomas) to be published by Cambridge University Press in May.