AUGMENTING MILITARY OPERATIONS USING AUTONOMOUS SYSTEMS
The U.S. requires defense systems that can do more with less human-intensive tasks. The next-generation of autonomous systems represent an improvement over today's capabilities, but are still too fragile for complex, uncertain, unstructured environments and complex missions. Bottom line: human involvement is still required with autonomous systems in order to deal with the unexpected.
The next level of autonomy requires systems that comprehend their environments and relevant aspects of the battlespace in the context of the commander's intent and objectives and, when necessary, in collaboration with human teammates.
The Department is pursuing the sustained research necessary to include: perception and situational awareness; adaptation and learning; and complex system dynamics. Interest in autonomy also extends to the collection and processing of data, to mitigate limits on the speed at which information can be absorbed and prepared for analysis. Systems that reason and mimic human cognitive capabilities, and systems that weigh different courses of actions based on knowledge and prediction are research endeavors that are increasingly showing promise in application to autonomous systems and will continue to be encouraged. To realize the full benefit of autonomy, the Department plans research in perception and situational awareness; adaptation and learning; trusted system; and complex system dynamics. Autonomy also requires that we address the problem of human-in-the-loop systems, to integrate automation with human comprehension and decision-making.
S&T Emphasis Area Resources
Autonomy S&T Roadmap