Intelligent Sprinkler: Design and Evaluation of a Low-Cost Mobile Agricultural Robot for Sensor-Driven Irrigation and Vision-Guided Targeted Spraying

Uniform irrigation and blanket spraying can apply water or treatment liquid without accounting for local field conditions. This work presents the design of the Intelligent Sprinkler, a low-cost tracked mobile agricultural platform that combines soil and climate sensing, sensor-driven irrigation, camera-based target detection, targeted liquid delivery, obstacle-aware navigation, and feedback logging within a SENSE–PERCEIVE–DECIDE–NAVIGATE–ACTUATE–MEASURE–FEEDBACK loop. The platform uses a Raspberry Pi for high-level processing and a microcontroller for real-time sensing and actuation. Irrigation is implemented as a threshold-based closed-loop controller rather than as a machine-learning irrigation predictor, while YOLOv8n is specified for target detection and edge deployment. The evaluation protocol defines controlled irrigation, vision, spraying, navigation, and complete-cycle experiments with predefined metrics and statistical tests. This manuscript includes an explicitly labeled synthetic dataset to demonstrate the intended reporting structure. The numerical values are not experimental findings and are placeholders for verified measurements from the physical prototype.