Data Fusion Research

Data fusion is generally defined as the use of techniques that combine data from multiple sources and gather that information in order to achieve inferences, which will be more efficient and potentially more accurate than if they were achieved by means of a single source.

iGSI performs data fusion research primarily in the context of the Sensor Web and the Semantic Sensor Web. The Sensor Web represents an infrastructure that allows multi sensor data fusion. This process is highly complex and includes the aspects information aging, i.e. the homogenization of incoming data sets and their extrapolation to specific points in time, information weighting of individual sensors or sensor readings, and information interpretation. The latter usually requires the transformation of unstructured, initially incommensurable data structures and information models into a structured representation. This process includes a semantic enrichment of the data, a process that gets further elaborated in what is called the Semantic Sensor Web.

iGSI investigates the potential of the Semantic Sensor Web to complement human sensing of the environment. Humans are often highly efficient at filtering, contextualizing, and discriminating data. Humans are usually better than machines to adapt their activities based on the latest sensing and perception. This advantage yields from the capability to leverage extensive background knowledge in combination with highly complex reasoning mechanisms and a virtually endless set of experiences. iGSI performs research on how those human capacities can be matched or complemented by computer and sensor networks with the goal to optimize situation specific decision making. Human deductive powers get complemented by computer symbolic processing power to enhance situational awareness and understanding.