AI-driven Cognitive Intelligence for Defense Applications

One of the potential contributions of AI to defense applications is the ability of machine learning to drive complex decision-making processes beyond directed autonomy. This ability is broadly captured under cognitive intelligence in which the agents, by design, develop intelligence to drive observe, orient, decide, act (OODA) loop. As such, current research is geared towards further transforming this intelligence towards natural interaction across agents and between agents and the decision-makers. In the lecture, we will provide two AI-driven design methodologies in which such cognitive intelligence is developed for Unmanned Surface Vehicles (USVs) and Unmanned Aerial Vehicles (UAVs). First example showcases an AI-Aided tactics generator for USVs to pursue, deter and neutralize a diverse array of security threats, from piracy and smuggling to potential terror attacks, thereby providing maritime infrastructure protection. Second example showcases a natural language model coordination between a swarm of UAVs and operators. This model interprets high-level commands, facilitating communication at a strategic level. The implementation of Intelligent Mission Planning (IMP) capabilities reduces the need for intricate human-in-the-loop commands. This natural command-control interaction leads to high-level mission effectiveness as demonstrated on a typical reconnaissance and target acquisition mission conducted by a swarm of UAVs.