ABSTRACT
In order to provide a basis for the development of autonomous sys-tems architectures capable of supporting ethical behavior regarding the application of lethality in war, we now consider the use of mathematical formalization as a means to express rst the underlying ow of control in the architecture itself and then how an ethical component can eectively interact with that ow. is approach is derived from the formal methods used to describe behavior-based robotic control as discussed in [Arkin 98] and that has been used to provide direct architectural implementations for a broad range of autonomous systems, including military applications (e.g., [MacKenzie et al. 97, Balch and Arkin 98, Arkin et al. 99, Collins et al. 00, Wagner and Arkin 04]).
