Engineering for Emergence in Information Fusion Systems: A Review of Some Challenges

Modern Information Fusion (IF) systems are faced with evolving operational environments where human and intelligent systems will function as a team to achieve mission objectives. These evolving operational contexts demand a full spectrum dynamic response of 'data-to-decision' from IF systems. Traditional information extraction and fusion levels typically address the "data" end of the spectrum, while recent advancement in Machine Learning (ML) and Artificial Intelligence (AI) approaches are being used for the "decisions" end of the spectrum. However, the IF system behavior emerges from the various complex interactions that take place between different fusion levels (including human interaction), the operational context, and the employed AI/ML techniques. In this paper, we explore this emergent behavior of the IF system and argue that holistic system design and evaluation techniques, as offered by System Engineering (SE), provide means to recognize and characterize this emergent behavior. Furthermore, we describe the research challenges for future IF systems that will enable managing emergence by leveraging SE, while exploiting the context-aware information fusion aided by the advancements in AI/ML.