| Popis: |
The “industry of the future” highlights technological transformation via the digital revolution as a major axis for building future industrial performance, promising flexibility and improved working conditions. Many of these technologies aim to be autonomous. We will question this desire for machine autonomy and more especially the principles on which it is based.We present the main results of three autonomous device design projects: the introduction of an industrial cobot, the design of an autonomous agricultural robot, and the design of autonomous shuttles. For each of these three projects, we will demonstrate the limits of autonomy, when it means leaving aside the design of the relationships between the entities, including those that are autonomous, in the future system.By avoiding these issues, the designers reduce the complexity of their task. But the contextual conditions that favour the effectiveness of the performances targeted by the introduction of the technology are then often missing. In such conditions, the human operator remains the main adjustment variable through which integration, albeit degraded, remains possible.In line with previous proposals in activity ergonomics, we defend the importance of addressing the question of the relationship very early on when designing devices that integrate autonomous technology. To this end we defend the utility of a design approach based on the extension of action capacities, as well as the utility of paying greater attention to the political dimension of projects and thus to the vision of future work that these emerging technologies implicitly support. |
