Consider the following situation. A large network connects a considerable number of robotic vehicles and automatic observation stations, together with a number of human operators that wish to use the information from the robots and observers, and to assign tasks for the robots to perform. Operators do not communicate directly with the robots. Instead, each operator uses an operator's assistant, a computer system that communicates on behalf of the operator with the various robots as well as with the other operator's assistants in the network. Moreover, these operator's assistants are autonomous in the sense that they are able to perform some operations of their own, so they are not merely a communications interface.
Different operator's assistants may take on different roles. In particular, some of them may have the role of coordinating the activities of others.
The research in the Laboratory for Cognitive Autonomous Systems has the design of such autonomous operator's assistants (AOA) as its primary focus at present, using techniques from artificial intelligence and in particular from knowledge representation. We participate in the WITAS project by developing an AOA system that mediates between a human operator and an autonomous UAV (uninhabited aerial vehicle).
Since robots and AOA's share the property of autonomy, it is reasonable to assume a certain degree of similarity in their respective designs, and also some significant differences. The following capabilities are characteristic of cognitive autonomous systems of any kind: