Exploring technology and the potential to bridge the gap between current technology and its development and application to the future battlefield

While processing of large volumes of distributed sensor information has been demonstrated, it is only amenable to direct AI processing, and not for interpretation by a human.
Control of assets is also assumed to be instantaneous and centralised, which is infeasible for real world application. There is no emphasis on collaboration with humans, either in cooperating with humans as part of a mixed Human-AI team, or the AI providing any suitable mechanisms for explaining its actions or intentions to a relevant operator. There exists a technology gap between current SOTA, and deployable AI for the future battlefield. It is addressing questions in this technology gap that interests me, and the area that I would like to research.