The combination Computer Based Training systems with Artificial Intelligence and Cognitive Science has led to the development of Intelligent Tutoring Systems nearly 30 years ago. A common agreement has been reached about the constituents of an Intelligent Tutoring System (ITS). Nonetheless, the interpretation of the role of each component in the ITS is still heterogeneous. Together with the absence of formal methods in ITS, this leads to the situation, that the components of ITSs are strongly domain dependent and not reusable. The situation is analyzed for case-based ITS, i.e. ITS based on a narrative story line. A formal model of the tutoring process is introduced into the ITS architecture. The model is based on the idea to support the training of two cognitive processes, i.e. the process of diagnostic reasoning, and the process of general knowledge application. The integration of the tutoring process model as the central component in the ITS has led to a homogenization of the archit...