As inexpensive computers add limited intelligence to a wider variety of everyday products, a new model of computing becomes possible.
This new technology aims for the flexibility of a far simpler and more ubiquitous technology: printed text. Depending on the need, print can be large or small, trivial or profound, verbose or concise. But though print surrounds us in myriad forms, it does not dominate our thoughts the way computers do today. We do not need to log on to road signs to use them or turn away from our colleagues to jot notes on a pad of paper. Similarly, ubiquitous computers would demand less of our concentration than present commercial computer interfaces that require users to sit still and focus their attention. Yet through casual interaction they would provide us with more information and all the advantages of an intelligently orchestrated and highly connected computer system.
Creating such an intuitive and distributed system requires two key ingredients: communication and context. Communication allows system components to share information about their status, the user and the environment---that is, the context in which they are operating. Specifically, context information might include such elements as:
The combination of mobile computing and context communications can be a powerful one [40,30,28,32,33,31]. Consider, for example, an employee who wants to show a set of figures to his manager. As he approaches her office, a quick glance at his tab confirms that the boss is in and alone. In the midst of their conversation, the employee uses the tab to locate the data file on the network server and to request a printout. The system sends his request by default to the closest printer and notifies him when the job is finished. Many more examples of the Ubiquitous Computing philosophy are presented in Mark Weiser's article ``The Computer of the 21st Century'' [39].
Attaining the goals of Ubiquitous Computing will require a highly sophisticated infrastructure. In the ideal system, a real-time tracking mechanism will derive the locations and operational status of many system components and will use that context to deliver messages more intelligently. Users will be able to choose from among a variety of devices to gain mobile, high-bandwidth access to data and computational resources anywhere on the network. These devices will be intuitive, attractive and responsive. They will automatically adapt their behavior to suit the current user and context.
Although one can speculate about the design of a future system, unfortunately the components needed to build such an infrastructure have yet to be invented. Current processors and microcontrollers are slow and power-hungry compared to their likely descendants 10 years from now. We reasoned that we could bridge this technology gap by constructing an operational system that resembles an optimal design. Despite the inevitable compromise of some engineering characteristics, we could then use the system to assess the advantages and disadvantages of Ubiquitous Computing as if we had glimpsed into the future.
It is impossible to predict the range of device forms and capabilities that will be available a decade from now. We therefore based our device research on size, a factor that is likely to continue to divide computers into functional categories. A useful metaphor that highlights our approach is to consider the traditional English units of length: the inch, foot and yard. These units evolved because they represent three significantly different scales of use from a human perspective.
Researchers at PARC have built computer systems at the three scales described above [41]:
These experimental devices use different mechanisms for communication and computation within the building's distributed system. The Liveboard is not mobile and connects directly to an Ethernet. Our mobile devices extend battery life by using low-power communication technologies: infrared (IR) signalling for the PARCTAB and near-field radio [8] for the PARCPAD . We have also investigated how operating system design can reduce power consumption [43] and this is well suited to mobile computers. The PARCPAD and Liveboard are described elsewhere by Kantarjiev [17,12] and Elrod [9].
Our goals for the PARCTAB project were: