Collaborative Services Using Local and Personal Facts

Ubiquitous Computing is starting to be reality. The office and home environment of tomorrow has a much higher density of computing devices. Presumably a lot of these devices will be public in the sense that they can share their services with other services or users. Some of the devices will also be wireless and mobile. This dynamic service space will require a new information infrastructure for distributed computing environments, to cope with the increasing complexity. We introduce the notion of Local Facts and Personal Facts, as key components for enabling context sensitive services. These facts are provided trough information repositories and make it easier for services to find relevant services and facts. We also introduce a Gadget Memory as an information repository for Local Facts, and a Personal Agent as a repository for Personal Facts. COLLABORATIVE SERVICES "In the morning your alarm clock starts to ring. After a while you get into your slippers. At that point the coffee machine starts to make your morning coffee. A couple of hours earlier you did put on the slippers when you went to the toilet, but that didn't start the coffee machine." The computing environment of the future will (if the trend continues) consist of small devices providing a limited set of specialized services, e.g. a slipper that can provide the information whether it contains a foot or not. Looking at each device individually, the type of services they can provide seem kind of limited and unnecessary, but in combination they could have enormous potentials. The service that helps you get a fresh cup coffee in the morning involves a number of different devices. The multi-device service uses the characteristics of each device, to provide a useful service. So how do we get the different services to collaborate? One basic property that must be enabled is the ability for the different devices to communicate. Radio communication such as Bluetooth or WLAN technologies can provide the low level data package communication, while software agent technology provides some interesting features such as higher level communication protocols and support for negotiation. A crucial issue for collaborative services is to know with which other services it is interesting to collaborate. Assuming that the devices are connected to the Internet, there could be an infinite number of possible collaborators. The major part of these services are not interesting though, even if they provide the type of service that we are looking for. The coffee machine in the example above would for example not have much use for the information that someone put a foot in a slipper on the other side of the world, (or next door for that part). There is a need for infrastructure to support generic working ubiquitous computing systems [5,12]. Important for the infrastructure of such systems is context relative information about devices, services and their users. Instead of using a flat address space, such as Internet addresses, where all connected devices are equally addressable, context relative information could be obtained from informationor fact repositories, containing only information with certain properties. One example of such a repository is the Jini Lookup Server [15], containing information of services available on a local network. In this paper we will examine two other types of repositories: The Gadget Memory, handling only local facts describing the computing biotope, that is, facts that are only relevant to a physically limited area like a room or a flat. The other repository is the Personal Agent, handling personal facts, that is, information related to a certain person, e.g. information about personal preferences or contact information. THE PUBLIC AND PERSONAL COMPUTING DOMAINS When interacting with computers today, we move in different computing domains. We have a personal domain consisting of our personal computing devices, as well as of data and services that are in some way related to us. There is also a public domain, consisting of computing devices, services and data open for anyone to use. The public domain can be divided into a local environment, the computing environment in your nearby physical environment, and a global environment consisting of Internet services and information without local or personal connections. Information Services Personal