Autonomous Systems: Issues for Defence Policymakers
This chapter recommends the increased use of agent-based simulation methods to support the design, development, testing, and operational use of autonomous systems. This recommendation is motivated by deriving taxonomies for intelligent software agents and autonomous robotic systems from the public literature, which shows their similarity: intelligent software agents can be interpreted as the virtual counterparts of autonomous robotic systems. This leads to examples of how simulation can be used to significantly improve autonomous system research and development in selected use cases. The chapter closes with observations on the operational effects of possible emergent behaviour and the need to align the research agenda with other relevant organisations facing similar challenges.
Original Publication Citation
Tolk, A. (2015). Merging two worlds: Agent-based simulation methods for autonomous systems. In A. P. Williams & P. D. Scharre (Eds.), Autonomous Systems: Issues for Defence Policymakers (pp. 291-317). Norfolk, VA: North Atlantic Treaty Organisation Headquarters Supreme Allied Commander Transformation.
0000-0002-4201-8757 (Tolk, Andreas)
Tolk, Andreas, "Merging Two Worlds: Agent-Based Simulation Methods for Autonomous Systems" (2015). Modeling, Simulation & Visualization Engineering Faculty Publications. 32.