Since AI-enabled autonomy learns by prior training (often via learning-enabled components or LECs) and through experience, it adapts its actions to changing scenarios. An AI system processes a large amount of data in a short time, finding the best solution available at that time for the task at hand for the autonomous system or human-machine team. It can also perceive, assess, and make sense of new situations given very little data – also known as “sparse data” – which shortens decision-making timelines.
When AI-enabled Autonomy must meet a higher standard
This ability to interpret information and make good decisions quickly in new situations is useful for any autonomous technology, whether it is used in civilian, industrial, or military environments. However, the challenges and potential consequences of those abilities are vital when in use by defense and intelligence agencies. An autonomous military system (AMS), for example, not only will often need to perform complex tasks in disrupted, chaotic, or unanticipated environments, but may do so while being attacked by adversaries. Also, of course, the stakes are always higher when a deadly force is involved. As a result, AI-enabled autonomy for military use is held to a higher standard in most every measure, including situation analysis, decision-making speed, task capabilities, and response/behavior suitability.
While commercial and academic developers have made real progress in autonomy, in the U.S. the engineering expertise, accelerated innovation, and in-depth resources required to make AI-enabled Autonomy technologies dependable, adaptable, and trustworthy enough for military use is found mostly in federal research and development agencies. This includes the Defense Advanced Research Projects Agency (DARPA) and their partners in the defense community.
This information page is provided as a service to our readers by BAE Systems, Inc., a U.S.-based world leader in aerospace, defense, power, and intelligence solutions. Learn more about us here.
