For many years, the development of renewable energies and specifically the production on wind energy remains a challenge for the energy transition. A major effort has been made by manufacturers to offer a high availability (> 97) of their equipment whatever the environmental conditions. However, according on the chosen sites (wind conditions favorable or not), the profitability of equipment may be low (around 1 %). In order to improve the cost control over the lifecycle of the equipment, it is then necessary to identify the process of degradation of the equipment. Therefore, to respond to this problem and to help wind farms to improve their maintenance policy, we present the development of a decision support tool for the planning of just-in-time maintenance activities based on the triplet Quality-Cost-Delay (Just-in-Time Monitored Maintenance). Within the framework of this paper, we present a state of the art of the diagnosis and prognosis of the health state of technical systems and we introduce the goal of our study: the degradation of wind turbine gearbox. This component is more complex (two planetary gear trains) and its degradation appears to be quite variable overtime (average of 8 years with premature replacement components (5 years)). The replacement operations are difficult and costly and involve specialist personnel at 100 m of height. To limit the risk, the environmental conditions (sunshine, little wind) must be perfect and the interventions are planned several weeks or even months in advance. It is therefore necessary to build up our expertise on the characterization of the health state of this type of equipment. In order to identify the degradation and to anticipate premature wear by preventive maintenance on ?sensitive? equipment, we present the various modules developed: monitoring module, diagnosis module, prognosis module whose objective is the determination of the Remaining Useful Life of the system (RUL) taking into account the operational constraints. The monitoring module is based on the laws of conservation of energy for design analytical redundancy relation. Indeed, the temporal repetition of the values of certain variables constitutes an information redundancy whose coherence must be checked. His incoherence generates then an indicator (residue) defining itself as the abnormal operation of the system. The diagnosis module is based on a matrix of recognition of causes. This allows, from abnormal operation, to identify the likely cause of degradation. The prognosis module presents different predictive models of health indicators and allows estimating the remaining useful life of the system. However, depending on the different assumption chosen, it is therefore necessary to suggest various planning strategies of intervention to the operator who is the guarantor of the implementation of the strategy. This work concludes with various simulations using data from wind turbine farms. These allow the evaluation of the performances of the various modules.