Armed Forces must constantly be ready for deployment in both national and international operations. Large-scale exercises ensure that different forces learn to co-operate, work together as a team, and maintain a level of preparedness. There are however limitations to the level on which exercises can be performed. Lack of available personnel, material, specific capacities and insufficient space and budgets result in difficulties with performing exercises with higher echelons involved. Even for small-scale training events similar difficulties can occur, especially when team collaboration is required. To be cost-effective and simultaneously meet training goals, training often requires a mix of live, virtual, and constructive simulations (LVC). Synthetic wrapping is a form of LVC simulation that enriches training through virtual and constructive simulation. Simulated units that participate autonomously in live exercises and training modules need to be able to communicate with the primary training audience, behave perceptually realistic, and contribute to the learning goal.
This research aims to explore a variety of Artificial Intelligence techniques in the context of creating perceptually realistic synthetically wrapped units. It focuses on three different domains: (i) communication via natural language, (ii) understanding the environment, and (iii) adaptive behavior. We present a synthetic-wrapping framework, and we show it working on a representative use-case: Joint Terminal Attack Controller (JTAC) training. In the use-case, the JTAC requests Close Air Support (CAS) from a simulated fighter pilot and guides the pilot towards a target. The proposed framework successfully supports the use-case and supports adaptive behavior based on the environment and natural interaction with a simulated unit. Our results show that modern Artificial Intelligence techniques can aid in the creation of perceptually realistic simulated units that allow for believable interactions and behaviors during military training and exercises.
Keywords
Additional Keywords
SYNTHETIC WRAPPING, ARTIFICIAL INTELLIGENCE, MILITARY EXERCISE, SIMULATION, MILITARY TRAINING