This area develops theoretical foundations, models, and methods for resiliency of tactical battlefield analytics in dynamic adversarial settings. IoBT has extended the field of robust learning to scalable, decentralized, and dynamic solutions. IoBT solutions adapt to dynamic (non-stationary) environments. They advance the Pareto frontier of policies that balance performance (obtaining an accurate and timely situational awareness) and resilience to various adversarial attacks that manipulate data. The goal is to realize a distributed automated system that can manage and control resources in an sensor-to-effect/decision loop operating in a contested environment with ever-changing sensor phenomenology due to the fog of war and adversarial deception. The automated system needs to be able to control its decision cadence, quickly take confident actions when possible, and slow down and recalibrate as the environment changes to ensure all actions are safe.