This thesis presents an agent system, which supports a user in situation analysis and tactical decision making in a military context. The main issues of this work are the conception of the agent, the implementation of the agent in an application domain, and the evaluation in human-machine experiments. The core of the conception of the agent is the creation of a machine situation awareness. For this purpose, the agent processes the situation knowledge of an environment and integrates it into an internal model based on influence maps. Based on this knowledge, the agent applies algorithms for tactical decision-making to generate appropriate courses of action for specific situations that arise. These courses of action are communicated to the user in the form of recommendations. The generation of suggestions for the tactical approach is adapted to the situation and the user has the possibility to edit or modify the suggestions via system interactions and/or to make requests to the agent. Furthermore, the work includes the implementation of the concept for the domain of Manned-Unmanned Teaming, as well as the integration in a research simulator for transport helicopter missions at the Institute of Flight Systems at the University of the Bundeswehr Munich. Finally, the implemented agent system is evaluated in human-machine experiments in which subjects completed dynamic mission scenarios, among others. The results show that an intelligent agent for tactical support in military missions increases the effectiveness of the available means of operation and enhances the users' sense of security. The results also reveal potential for improvements and further research directions. Overall, the chosen approach of an intelligent agent system to support situation analysis and tactical decision making is very well suited.
«This thesis presents an agent system, which supports a user in situation analysis and tactical decision making in a military context. The main issues of this work are the conception of the agent, the implementation of the agent in an application domain, and the evaluation in human-machine experiments. The core of the conception of the agent is the creation of a machine situation awareness. For this purpose, the agent processes the situation knowledge of an environment and integrates it into an i...
»