Modelling and Reasoning with Imperfections in Situation Monitoring
Background
Situation awareness involves the real-time processing of event-based information coming from an evolving situation in an attempt to understand what is happening. It relies mainly on the determination of environmental parameters through devices like sensors. A crucial problem associated with such devices is their inherent imperfections which affects significantly assessing the situation, its risks or opportunities in situation awareness. For instance, some information needed may be imprecise, uncertain, incorrect, or not available, or may origin from sensor that cannot be fully trusted / reliable. Further, information from one sensor may be conflicting with information from another sensor or with the existing knowledge. By such imperfections, an exact (precise and valid) situation assessment as needed for a decision may not be possible, though it may still provide useful information by constraining the situation space and thereby decision space. Research in this area poses a significant challenge and is very much limited so far.
Research Program
To overcome this deficiency and treat imperfections common in sensor data, this proposed research will propose an extension of context representation, modelling and reasoning with imperfect information in cases of situation awareness. The emphasis of the proposed research is on utilizing statistical techniques in combination with information fusion, fuzzy logic, case-based reasoning and decision making approaches to accommodate the imperfect nature of the captured context and to act as a basis for developing methods for modelling and reasoning under imperfection. The existing and current research works by supervisory team in these areas form the basis for this project. Finally the project aims to develop a context-reasoning engine taking into account a range of aspects of imperfection in situation awareness.
Anticipated Outcomes
The main domain of application is Smart Homes for Assistive Living. The developed engine will be evaluated through a series of experiments by applying developments to data collected in the Smart Lab from the Smart Environments Research Group (SERG) in School of Computing and Mathematics, University of Ulster.
Personnel Involved
First Supervisor: Liu, J Dr
Second Supervisor: Chen, L Dr
Third Supervisor: Nugent, CD Professor
Collaborator: Prof Luis Martinez-Lopez (Spain)
Collaboration: This project does not involve collaboration with another establishment
Synopsis:
Situation awareness involves the real-time processing of event-based information coming from an evolving situation in an attempt to understand what is happening. It relies mainly on the determination of environmental parameters through devices like sensors. Research in this area poses a significant challenge and is very much limited so far in terms of handling imperfections common in sensor data. The emphasis of the proposed research is on utilizing statistical techniques in combination with information fusion, fuzzy logic, case-based reasoning and decision making approaches to accommodate the imperfect nature of the captured context and to act as a basis for developing methods for modelling and reasoning under imperfection. It is anticipated that a successful completion of the project will result in a new context-reasoning engine taking into account a range of aspects of imperfection in situation awareness with application in Smart Homes for Assistive Living.