Production Activity Control, considered as the operational levels of production management, requires quick decision making in order to react to changes on the resources or on the workshop environment. The design of a Decision Support System in such a context implies to model the reasoning of the experts of the field. A simple conceptual model of Production Activity Control is suggested, in order to identify the types of reasoning used. The uncertainty and imprecision of the information led us to use fuzzy logic in order to select various inference semantics, adapted to the types of decision identified. Examples of implemented reasonings are described.