Autonomous battalion simulation for training and planning integrated with a command and control information system

Current Command and Staff training uses simulation systems that consist of computer generated forces in combination with human operators. The human operators receive high level tasks (e.g. company level) as input, transform these into lower level tasking for subordinate units (platoon level and lower), and then they manually enter the more detailed sets of instructions into the simulation system. A challenge in the case of training is that the amount of resources required inhibits a high frequency of training events. The need for a large simulation supporting staff is even more problematic if simulations were to be used more during operations in planning or mission rehearsal, e.g. for what-if analysis. We are investigating how we can make a more autonomous simulation system, which interfaces Command and Control Information Systems (C2ISs) in a seamless way, minimizing the number of human operators. In order to realize a seamless integration of a simulation system with a C2IS, an order made in the C2IS must be expressed in a standard, unambiguous language, which is interpretable by the simulation system. Also, to make the simulation system able to carry out higher level operations (e.g. battalion operations) autonomously, the simulated forces must have sufficient knowledge about tactics and doctrine. In this report we describe the design and implementation of a first version of a demonstrator of a simulation system capable of autonomous simulation of battalion operations. The simulation system is integrated with a C2IS, which can be used to visually view and create orders, in addition to presenting ground truth and perceived truth. The simulation system is capable of receiving and executing orders created by the C2IS and providing reports back to the C2IS. In addition, the C2IS is used to define the ORder of BATle (ORBAT) and provide the initial positions for friendly units. The exchange of orders, reports and scenario definitions between the C2IS and the simulation system are expressed in Coalition Battle Management Language (C-BML) and Military Scenario Definition Language (MSDL), where C-BML is a standard under development for exchanging orders interpretable by machines and MSDL is a standard language for describing scenarios. The simulation system consists of a Multi-Agent System (MAS) and a commercial off the shelf CGF system. Knowledge of higher level tactics and doctrine are implemented in the MAS, which is used to control the entities in the CGF system. The agents in the MAS are organized in a hierarchy and represent leaders and staff of military units. The behaviour model is based on the human behaviour modelling paradigm Context-Based Reasoning (CxBR). The implementation of the CxBR based MAS framework and the behaviour model implementation are documented in this report. The simulation system integrated with a C2IS was demonstrated for subject matter experts (SMEs), and the response was generally positive. Feedback from SMEs, regarding possible applications for such a system, what the system should and should not do etc., is summarized in the report.