This research is the result of four years of practical and scientific investigation of the phenomenon of traffic evaporation, which was considered and then demonstrated to be the opposite of traffic induction. It has anchored, in practice and in time, an orphaned and scarcely documented topic in literature which mobility planners are facing. Our research was conducted with one foot firmly grounded in the practice of mobility engineering and the other in the urban sociology of mobility behaviour. Immersed in the first field as a "project engineer" on various projects, I was able to investigate a great number of case studies and enrich the thesis with the questions of experts and commissioners. In order to answer these questions, beyond the quantitative and measurable aspects that account for the engineer's daily bread, the qualitative methods of urban sociology were mobilised to decipher the rationales for action of these individuals that make up traffic flows. This singular approach revealed the number and diversity of factors that are at play in the mechanism of traffic evaporation. These include motility, the mental cost of learning, the anchoring of modal habits and the value rationale of each individual. At the crossroads of the three research foci, which are usually treated separately and which we have combined in our case studies, i.e. the temporality of disruption, the (re)actions of the authorities and the weight of (multi)modal habits, we have been able to observe orders of magnitude beyond traffic evaporation. The results show, first of all, that traffic evaporation is achievable over the long-term of urban planning, because the time of mobility habits and learning of usage is also long, although it seems to accelerate once learning is triggered. Moreover, our results show that traffic volumes which we thought would always increase in the past can be controlled. This legitimate confidence in the reversibility of traffic induction draws attention to levers that are now, in turn, perhaps being underestimated in mobility planning. Secondly, the study of urban construction sites has shown that the keys to understanding the reactions of the traffic flow lie in the varied profiles of the individuals who make it up. Knowing their rationales for action, it is possible to assess which behavioural changes will be temporary and which will be lasting, and thus to steer them if desired. Finally, we have revealed that, in the face of the unforeseen, the activation of motility levers is strong and prompt, with the aim of overcoming personal barriers to change. In these cases, the behaviours adopted seem to last beyond the duration of the disruption. Thus, traffic evaporation appears to be much more than a new variable to be considered in mobility analyses and planning. It calls into question the logic that has driven travel planning in the automobile era.