The scientific domain of our research work is Multi-Agent System. In this domain, we are interested in the development of global models defining and structuring the common activity of autonomous agents. The autonomy of an agent results in its capacity to determine its own goals vis-a-vis the configuration of the environment and its motivations. Because of this autonomy, conflicts related to the existence of goals suitable for each autonomous agent of a society can be incompatible, complementary or in competition with the society objectives. Models of conflicts management exist and within the framework of this work, we focused on the organisational models. We assert that the definition and implementation of global constraints influencing autonomous agents functioning must be done so as to take into account the possibility that the agents can not respect constraints according to individual contexts and objectives' of each one. In order to achieve these goals, we propose an Electronic Institution model composed of a constraints' description and specification model as well as a system of supervision and reinforcement of these constraints. This model is used to define and structure the activity of a set of autonomous agents within a normed organisation thanks to MoiseInst and to supervise the respect of the organisation specification with Synai. This arbitration system is composed of a set of supervisor agents able to implement constraints' regulation and reinforcement mechanisms on the autonomous agents. The model experimentation is done in two completely different domains. They concern the contents creation and execution for interactive television on a side and the management of electronic contracts in a e-commerce platform on the other side. They have the same arbitration of autonomous entities having to respect constraints issue. Moreover, both applications are complementary by their Institution scope.
