MOP-N: A Hybrid AI Model for Automated Deployment of Guns in Dynamic Wargaming Scenario

Abstract

Wargames play a critical role in training of military officers/commanders for operations planning and execution. A typical training wargame may involve a large number of players as part of more than one competing forces. It is often desirable to reduce number of players to improve playability of wargames, which is in conflict with the requirements of operational realism. Successful automation of players and their decision making processes can result in reduction of number of players without loss in the operational realism. Also, depending upon the objective, a wargame can be played at various levels of abstraction viz. procedural, tactical, operational and strategic. Usually higher level (operational and strategic) wargames involve abstract (low resolution) entities. Developing models to simulate these abstract entities is most critical challenge for wargaming simulation designers. Alternatively, with incorporation of automation of lower level players’ decision making processes, detailed high resolution entities can be represented enabling modelling at abstraction closer to real physical entities, which is relatively simpler. The automation may be required at different stages of a wargame. It may be required during game initialisation or in relatively static situations where time constraint on decision making (and thus computation time) may not be severe......