ACUMEN: Adaptive Course of action selection for Uncertain Mosaic warfare ENvironments

Recent conflicts around the world have demonstrated that the nature of the modern battlefield is significantly more complex than the battlefield of just 5 years ago. Between the data coming from the internet of battlefield things, extensive usage of UAVs for reconnaissance/kinetic capabilities, and overall increase in capabilities derived from combining assets across several service branches into a mosaic, the decision space in the field has exploded beyond its already large size. This evolution brings about a significant challenge: the overwhelming volume of data produced by sensors and intelligence gathering. The traditional methods of manual data inspections are no longer feasible, as analysts are inundated with vast amounts of raw data but struggle to extract actionable information. This phenomenon, often described as ‘drowning in data but starving for information’ is pervasive across multiple domains, including military, intelligence, healthcare and finance. Addressing this problem requires a scalable methodology for hypothesis management, capable of handling the exponential growth of decision spaces. By automating the evaluation of options and presenting the best alternative clearly and concisely, decision-makers can focus on the most plausible and impactful COA (Courses of Action). This capability is essential for maintaining situational awareness and making informed decisions in real-time. Such systems would partner with the human operators to enable them to reach decisions faster by organizing the COAs based on the most recent data. These metric estimations need to consider the dynamic set of battlefield assets provided by the mosaic of capabilities, complex engagement rules in the particular battle theater, the cost of enacting a COA and complex performance metrics interdependencies. Managing this decision space complexity in time scales short enough to be effective will be a key aspect of decision superiority in future campaigns.