Application of mobile agent systems to first responder training

In first responder training, flexibility is important. Applications that have flexibility provide great benefit for the trainee as they are able to try many things and have the training system respond appropriately and provide new scenarios to train on. The ability to change and create new scenarios is critical. However it sometimes requires a total rewrite of the application to handle all the new features. The ability to easily script out all of these changes and execute them dynamically and also originating from foreign sources would be very powerful. There are currently hardly any such applications that add this kind of flexibility and have the power and ease of use that is necessary. Mobile-FIRST, a mobile agent based system for first responder training, was developed in this thesis research by adding interpreted scripts and mobile agents in order to add much needed flexibility to training systems.;Using an embedded interpreter that can interface with the running binary application and a mobile agent system for mobile code agents to migrate and execute in these applications, a flexible and powerful mobile agent based system can be achieved. Design principles that needed to be included in the system were looked at as well as different architectures that can take advantage of a mobile agent based system. There are many issues involving a system such as this, such as the dynamic composition of tasks and implementation details integrating with a real-time system. The features of the system as well as the issues involved in using such a system are described in detail. A case study is then examined where we have integrated a mobile agent system into a first responder training video game. Adding this system allows for easier scripting of events in the game as well as being able to plug in and replace modules of the running game.