The Levels of Data Fusion

    Data fusion is the method of fusing data to create new data. Fusion means that there is a combination of two or more pieces of data to create a new piece of data. It is useful when you need to create new, meaningful data from previous data. In my Briefly Defining: Data Fusion post, I mentioned that data fusion has five primary uses, which were (1) pre-processing data, (2) object assessment, (3) situation assessment, (4) impact assessment and (5) decision making. We are going to look briefly at the levels of data fusion and what makes them different from each other. 

    The data fusion is split into five types of fusion, called the JDL levels of data fusion. The JDL is the Joint Directors of Laboratories, which was a group formed by the Department of Defense to specialized in the research of data fusion techniques back in the 1980's. The JDL created a level system to describe the different types of fusion methods and their uses of data in systems. This was to better understand the goals between the types of data fusion methods and the level of information they contribute to creating knowledge in a system. This post will briefly touch on each of the levels of the JDL data fusion. 

Note: It is important to be aware that the levels are not ordered by sequence of use, but are ordered by the knowledge they build and complexity. 

Level 0: Pre-processing data

This level fuses data to create a refined data set that provides useable data.

Level 1: Object Assessment

This level fuses data to create estimations, classifications and predications about objects. It creates information about the "what" and "where" components are.

Level 2: Situation Assessment

This level fuses data to create an understanding of a situation, event or behavior. It creates information about the relationship between objects. This level takes low level information, such as numbers, and converts it into high level information, like ideas/concepts, that humans can understand. 

Level 3: Impact Assessment

This level fuses data to create an understanding of the outcomes of a situation. It evaluates the different paths and its consequences for the future. 

Level 4: Process Reinforcement

This level fuses data to determine decisions to make and evaluating the reward(s) of the decision taken by the system. The system can use this knowledge to refine or tune its algorithm to make future decisions to reach the maximin or desired outcome.