Context Modeling in OWL for Smart Building Services

François Bry, Takashi Hattori, Kaoru Hiramatsu, Takeshi Okadome, Christoph Wieser, Tatsumi Yamada
2005 Workshop Grundlagen von Datenbanken  
This article presents an approach to context-aware services for "smart buildings" based on Web and Semantic Web techniques. The services striven for are first described, then their realization using Web and Semantic Web services is explained. Finally, advantages of the approach are stressed. Introduction An automated recognition of the people's location in a building (at desk, in a room, underway, etc.) as kind of so-called "contexts" [Dey01] makes it possible to offer advanced services such as
more » ... 1. context-aware message delivery, 2. context-aware access control, 3. context-aware services for disabled persons, and 4. context-aware management of the technical infrastructure. Such context-aware services are referred to as "smart building" services. As a first working hypothesis it is assumed that data on room occupancy, people's locations, and people's activities (e.g. working at one's own desk using a desktop computer, sitting at a conference table, etc.) are collected using cheap sensors such as RFID sensors. Especially RFID sensors can be placed at various locations such as in a gangway, on a chair, etc. The restriction to cheap sensors such as RFID sensors, excluding, e.g., video cameras, is convenient because such sensors are easy to install and to maintain, and processing such sensor data is simpler than, e.g., image processing. As a second working hypothesis it is assumed that sensor data is expressed in Semantic Web formalisms such as OWL [W3C04]. A Semantic Web approach to the "smart building" perspective is convenient because the Web provides with an widespread and easily deployed data interchange infrastructure, and the emerging Semantic Web provides with an increasing number of versatile data modeling and reasoning methods. This article describes an approach to such "smart building" services relying on the Web and on Semantic Web formalisms and methods. Central software components of this approach are described: The modeling of knowledge is realized in different ontologies called Micro-Theories written in OWL. The Micro-Theories are used to specify domain specific knowledge, e.g., of the furniture of a building. Such ontologies are integrated into one "Upper Ontology" specifying common sense knowledge such as the ways how to use the furniture. The next component is the derivation of context. As last component, data retrieval is realized through simple queries and ontology reasoning using Semantic Web Query Languages. Finally, this article explains why standards (1) for the data, sensors might collect in buildings, (2) for an ontology of common life objects, properties and action, and (3) for an interface between the afore mentioned formalisms make it easier to reason on such data and knowledge. 2 Context-aware Services for a "Smart Building" An automated recognition of people's location (at desk, in a room, underway, etc.) in a building makes it possible to offer advanced services, especially: Context-aware message delivery. Nowadays one can communicate from every place in an office building, and additionally one can choose between various kinds of communication technologies. Written text or spoken voice can be transmitted synchronous or asynchronous via a mobile phone. The flexible availability of users leads to a bigger extend of communication, catching one's attention frequently. While the sender can choose the time and the technology for communication depending on his context, e.g., while walking or being in the office, the receiver cannot. Obviously, a smart delivery of messages depending on the receiver's context, would make communication more convenient. A manager attending a meeting wants to get messages only, if they are relevant either for him personally or for the meeting. Hence, instead of arriving during the meeting, a less important email could be delivered after while walking back to his or her office. Since reading email while walking is not convenient, the message could be delivered as spoken text automatically. Context-aware access control. Office buildings are not open to the public for security reasons. Hence, different strategies such as using different door keys provide access control. This technique causes especially two problems:
dblp:conf/gvd/BryHHOWY05 fatcat:wzycwtpwjzcmlk6pxfdvhjbyjq