Assured Autonomy programme launched

DARPA has launched a new programme called Assured Autonomy that aims to advance and enhance the adoption of autonomous systems in safety-critical defence applications.

The programme will work to improve the ways computing systems can learn and evolve to better manage variations in the environment and enhance the predictability of autonomous systems such as UGVs and UAS.

Despite technology advances that have led to the proliferation of unmanned systems, there remains a question about the degree of trust that can be put into these systems. Assuring systems operate safely and perform as expected is integral to trust, especially in a military context, and systems must also be designed so that operators can determine whether, once deployed, they are operating reliably, and, if not, that appropriate action can be taken.

Assured Autonomy aims to establish trustworthiness at the design stage and incorporate sufficient capabilities so that inevitable variations in operational trustworthiness can be measured and addressed appropriately.

Approaches such as the idea of ‘equivalent levels of safety’, which is based on the idea that an autonomous system must be at least as safe as a comparable human-in-the-loop system that it replaces, have not proved a practical way to assure the safety of autonomous systems due to the prohibitive number of trials that must be carried out to test it.

The alternative being suggested is that a goal-oriented approach is better suited for systems that learn, evolve, and encounter operational variations. In the course of Assured Autonomy programme, researchers will aim to develop tools that provide foundational evidence that a system can satisfy explicitly stated functional and safety goals, resulting in a measure of assurance that can also evolve with the system.