This paper takes as a basis that developments in agricultural technologies, in particular agricultural equipment and machinery (automation and robotisation), are important for the long term analysis of environmental management. It is based on the observation that projections developed from social, institutional and economic variables but which assume constant, or even non-existent, technological development are soon overtaken. The paper draws on an exercise to develop a projection as part of training at the Ecole Nationale du Génie Rural, des Eaux et des Forêts (ENGREF - National Rural Engineering School for Water and Forests) together with experts from Cemagref. It suggests that four stages require special consideration : formulation of the scope of such an exercise in terms of content and methodological and operational considerations; design and organisation of the exercise ; identification of the main results ; identification of the direction and conditions for extending such an approach, in particular depending on difficulties encountered. The exercise revealed in particular the need to cross disciplinary and cultural boundaries in such an approach. It also allowed the definition of research directions and the delimitation of conditions in which this research could be carried out.