DORYLID SYSTEMS

The whole point of autonomy is to remove the need for human intervention. Yet the operational performance of a collective grows with both the number of assets and the sophistication of their capabilities, and that has a direct impact on two major aspects of human involvement: asset control and asset servicing. Collective Autonomy must therefore be designed around its users in a manner that navigates the tension between the need for reducing human intervention required per asset during operation, and giving users more meaningful authority over collective behavior.

This means more than providing supervision or an emergency override. Users must be able to understand and configure collective behavior, shape it to operational needs, and intervene at the appropriate level whenever they choose to do so — whether across the entire mission, within part of the collective, or at the level of an individual asset. Exercising that authority should not require abandoning autonomy elsewhere in the system, but on the contrary dynamically reshape how autonomy is expressed at system-behavioral level.

The objective is neither to keep humans continuously in the loop, nor to lock them out of the finer details of their systems’ capabilities. It is to make their workload lighter while making their authority more effective whenever they exercise it.