Existing traditional agricultural systems consume an enormous amount of water and energy, with more than 69% of the world's freshwater volume attributed to agriculture. This fact leads to a degradation of water levels in several aquifers, as well as water quality. In Morocco, agriculture consumes about 88% of water resources, and the management of these resources is a significant issue, especially after the decline in rainfall rates in recent years. Designing an irrigation system that reduces water losses and evapotranspiration losses is the key to sustainable precision irrigation. This paper proposes an irrigation system based on an intelligent and efficient strategy for the automatic watering of agricultural areas and more adapted to the climatic conditions of Oriental Morocco. The system uses fuzzy logic to decide the duration of watering in an adaptive and precise way. Weather conditions (air temperature and solar radiation) and soil moisture (water content) are the inputs to the fuzzy system. The acquisition of these data is made by analog and digital sensors. The variables are fuzzified using trapezoidal and triangular membership functions. In the Fuzzification step, the Max-Min inference engine and the Mamdani rule base are used to decide the best decision for each given situation. The control strategy was implemented in the Arduino Uno R3 board. Typical three-day data from the summer and fall seasons were used to validate the proposed system. The simulation results showed that the proposed irrigation system maintains the soil moisture above the user-defined value and eliminates the under-irrigation risk.