Section: Research Program
Construction of context information
The term context here is understood in a very broad sense: it is any information to characterize the situation of an entity (a person, an object, a location etc.). A context is built using data captured in the environment. In pervasive systems, a typical approach is to build and maintain a logical representation of the real world (or model) from a mass of data captured in the environment, then analyze it to "find" the underlying context, for example the flow of traffic on a city scale, or human activities within a building, etc.
This approach poses different problems:
Considering these aspects, the TACOMA team focuses on two issues.
Localized aggregation and performance information from heterogeneous, unobtrusive, low-cost and unreliable sensors. The reactivity required in IoT systems and the aggregation of large amounts of data (and its processing) are antagonists. We study how to relocate the building of contexts near objects. The techniques covered use the principles of data fusion and are guided by different criteria: relevance of abstractions produced for pervasive applications (e.g., detecting user activity), anonymization of exploited raw data, processing time, and computing power required for processing.
The spatial layout of physical objects. The term layout can be understood in several ways: the co-location of multiple objects in the same space, the physical arrangement of two objects relative to each other (e.g., two objects arranged at right angles), or even the passage of an object of a physical area to another, etc. The context here is based on the physical layout of objects in space and their movements. In our approach, the use of "spatial" properties allows building pervasive applications in accordance with the principles set out in section 2.1 . These properties can be obtained in different ways: via electronic tags (RFID) for annotating the most common objects, by using specific properties for light waves or short-range radio, or even by using specialized sensors.