This paper focuses on the collective design and immediate execution of an agent-based model (ABM) by dynamically interpreting the activity diagrams of agent behaviours. To reach this objective, we have implemented an ABM of livestock producers facing drought conditions in Uruguay. The first step consists in implementing a standard ABM with pasture growth, herd dynamics and simple agents roughly imitating farmers' strategies. The second step is more participative since it consists in assessing the model with the real cattle farmers. As with most modelling processes, this evaluation phase requires feedback on model design. In order to make this assessment more lively and efficient, we have conceived a tool for drawing diagrams that can be immediately interpreted by the agents. Thanks to this new editor, the actors have quickly understood how the model worked and were able to criticize and modify it. Thus, this innovative modelling tool enables the involvement of stakeholders in co-designing ABM for participatory foresight studies. We hope it will facilitate the emergence of new and more efficient practices for farm management that can account for climate changes. (Résumé d'auteur)
展开▼
