Adaptation in Collaborative Virtual Environments for Training

Virtual environments for training, especially collaborative ones, require adaptability in order to be used in various situations. In this paper, we classify the needs for adaptation into two levels. The first one concerns the application setting. The second one is a dynamic adaptation, at runtime. The models proposed in this paper fulfill these two levels of requirements. Virtual world, virtual humans' behavior, scenario and application configuration can be adapted to a specific training situation. Moreover, an action selection mechanism enables actors to dynamically adapt their action scheduling to a situation. This situation is defined by a current context, past actions and a goal to achieve, namely a scenario. Finally we validate our models by presenting an illustrative scenario: the collaborative assembly of a piece of furniture.