Autonomous mobile robots are experiencing a growing demand in smart agriculture, as they contribute to environmentally responsible production practices in alignment with the Sustainable Development Goals of the United Nations. This paper presents a navigation strategy for autonomous mobile robots in sustainable agriculture. The proposed approach is based on a simultaneous localization and mapping algorithm that integrates multiple odometry sources, along with both global and local path planners capable of guiding the robot through vineyard rows. The navigation approach is tested using a Scout 2.0 mobile robot in both an open outdoor environment and a real vineyard. Experimental results demonstrate the performance of the robot in executing repeated paths and its robustness during an autonomous 600-meter survey through vineyard rows.