There is geometry in the humming of the strings, there is music in the spacing of the spheres.
To run a simulation training exercise, a scenario definition is required as a starting point as mentioned earlier. Typically, these include all the parameters for the simulators themselves such as entities used, time of day, weather effects, entity starting locations and munitions effects. In addition, a story (or more formally, a mission briefing) is given to the trainees given the similar information augmented with the mission description (e.g. deliberate attack, search for weapons cache, etc.). The unit leader then develops a mission plan that is shared with the trainee group as appropriate.
The notion of scenario generation can be generalized to other domains. For example, a cognitive rehabilitation scenario could include the task to be practiced, locations of items needed for the task, and possibly the layout of the training area itself. Ultimately, scenario generation is required for all training exercises to provide the context for the training to occur. However, it is a relatively expensive process and can benefit from automated processes. Performing the scenario generation process by hand is a very expensive proposition and causes a lack of scenarios for use in training. This can cause existing scenarios to be repetitively used, resulting in training with reduced effectiveness.
The IRL is performing an investigation of procedural modeling techniques to perform automatic scenario generation. Procedural modeling refers to a technique in computer graphics of using a set of rules to create models, textures and/or animations for a scene or part of a scene. It is used when creating the component manually would be cumbersome or expensive. Rather than storing the components necessary to create the component, procedural rules are stored that can be used to re-create the component.
We are building a system known as PYTHAGORAS (Procedural Yielding Techniques and Heuristics for Automated Generation of Objects within Related and Analogous Scenarios). This system is built on a plug-in architecture and allows us to explore different procedural techniques for automatic scenario generation. It also provides a mechanism for additional data for creating such scenarios (for example, trainee profile data may be available, which we can use to build a more effective scenario).