Education and training are identified as key elements to support the transition towards sustainable farming practices in France. Furthermore, the agricultural machinery sector is called to leverage this transition through closer consideration of the farmers’ needs. Hence, educational programs are expected to bridge the gap between agriculture and machinery design and innovation. In this paper, we will discuss the environmental conditions and context that led a group of students in agronomy at mastering robotics and how this could help improve educational programs on agricultural equipment. The context was a competition organized by a French applied agricultural research organization that challenged students to build a robot for the mechanical weeding of sugar beet. We present here the work of a group of master students composed of farmers’ sons, with a solid background in farming. The competition occurred in the second semester of their mastery of agricultural equipment and new technologies. Of note, during the first semester the students were involved in: (1) a project to define the specifications of an innovative tractor with students from an engineering school in mechanics, thus experiencing the interdisciplinary dialogue; (2) the design and production of a database of the agricultural robots available in France. The database led the students to define what a robot is, then to collect and organize technical information about robots already on the market or existing as prototypes. This provided them with an implicit catalogue of possible robotic solutions in agriculture. Nevertheless, the students considered themselves as complete outsiders for the robot competition because they received no specific training on robotics. Accordingly, they decided to follow strictly the competition specifications. The teachers adopted a design thinking approach, supporting the students only if needed, rather than promoting the students’ appropriation of the subject. The characterizing feature of the robot was the autonomous steering system: a torsion sensor mounted on a rubber strip. This simple and robust solution exploited the agronomic characteristics of the sugar beet that is sown on straight lines and is not damaged by physical contact. The farming experience of the students allowed them to choose the simplest sensor, which used the plants as physical boundaries of the robot navigation corridor. Furthermore, this choice was suited for working in field conditions and was easily understandable by the users. We will further discuss which environmental and educational conditions issued from this experience could help agricultural students at mastering new technologies and machinery innovation. FULLTEXT http://ifsa.boku.ac.at/cms/fileadmin/Proceeding2018/1_Rizzo.pdf