CADR-RAS Webinar: Harnessing AI: The Future of Efficient and Effective Defence Training

Michael J. Richardson, Mark Dras, Rachel Kallen, Patrick Nalepka, Gaurav Patil and Simon Hosking
Command-and-control (C2) operations form the foundation of military forces, requiring policies and structures that adapt to technological changes and ensure personnel can handle complex, high-pressure missions. C2 simulation and training are crucial for mission success but face challenges like high logistical costs and the need for flexible, scalable systems. Our research explores whether AI agents simulating expert-level, human-like performance can address these challenges and reduce the number of white and red-force personnel required in C2 exercises. In collaboration with CADR-RAS and the Australian Defence Science and Technology Group, we have been developing a interpretable computational framework using multimodal machine learning and generative AI to simulate expert operators who DECIDE on actions, COMMUNICATE critical information, and ACT in coordination with human and autonomous agents. Validated through C2-inspired multiplayer videogames, these models offer pathways for future Defence use in realistic C2 training scenarios and human-machine teaming applications.