The growing demand for sustainable agricultural practices emphasizes the need for innovative technologies that optimize crop management in controlled environments. This paper presents a comprehensive solution that enhances mobile IoT nodes for precision greenhouses, focusing on two critical advancements: autonomous navigation and computer vision for plant health analysis. The proposed navigation system integrates adaptive behavior, allowing the robot to execute user-defined commands such as specific numbers of greenhouse tours, time-specific activations, or periodic operations tailored to various agricultural scenarios. Additionally, the introduction of a dynamic sleep mode minimizes energy consumption during idle periods. Complementing these improvements, the research integrates a computer vision module designed to analyze the health of tomato plants. Leveraging deep learning algorithms, the system identifies and classifies common plant issues, such as yellowing, wilting, and infections, while simultaneously detecting and counting fruits. Using advanced object detection models, the system annotates anomalies and fruits with bounding boxes, enabling precise monitoring of plant health. Reports detailing plant-specific issues, including identified anomalies and associated plant IDs, are generated and transmitted in real time, providing farmers with actionable insights into optimize decision-making and mitigate risks. By combining adaptive navigation with advanced computer vision, this research introduces a holistic approach to greenhouse automation. The results underscore the potential of integrating intelligent robotics and AI to enhance productivity, sustainability, and decision-making in precision agriculture. These advancements pave the way for scalable, energy-efficient solutions that respond to the evolving needs of global food production systems.