Data Fusion means processes that collect and analyse all available data describing one and the same physical object, from information gathered by many different sensors and measuring methods. In Internet of Things systems and sensors, large amounts of data are generated, where objects and events are described by data in more than one way. In such cases Data Fusion is useful, as it increases the possibility of detecting and classifying than what is the case with a single sensor.
Savantic have, among other things, developed Data Fusion for a network of sensors used with trespassing alarms in the Stockholm Metro system. Thanks to Data Fusion, the system is able to tell real alarms from false ones in a more reliable way. The performance of this system surpasses existing solutions, which do not use different kinds of sensors and Data Fusion.