Breaking news in autonomy

Across the defense industry, autonomous systems are moving rapidly from demonstrations toward operational testing. The goal is not to remove humans from the battlefield, but to give them more capable machines to work with.

Autonomous aircraft can potentially navigate, identify objects, manage mission tasks and coordinate with other systems while requiring less continuous control from an operator.

That could fundamentally change the economics of air power. Instead of relying exclusively on a small number of extremely expensive aircraft, militaries could combine crewed fighters with larger numbers of autonomous systems designed for specific missions.

Companies including Shield AI, Anduril and Northrop Grumman are developing technologies aimed at making this concept practical. Their approaches differ, but the underlying challenge is the same: creating machines that can operate reliably when communications are limited, conditions are unpredictable and decisions have to be made quickly.

The technology is still developing. Questions surrounding trust, human control and battlefield reliability remain just as important as the underlying AI.

The next test

Autonomy will ultimately have to prove itself outside the laboratory. The systems that succeed will be the ones capable of working alongside humans when the environment becomes uncertain.