Agrova: AI-powered precision agriculture rover for intelligent crop management

Precision agriculture harnesses cutting-edge technologies to optimize crop management by addressing both spatial and temporal variability in fields. This paper introduces Agrova, an AI-powered autonomous rover purpose-built for intelligent crop monitoring in small-scale farms. Agrova features a rugged, custom-engineered six-wheel chassis—meticulously modeled in SolidWorks and structurally validated using ANSYS—to navigate diverse agricultural terrains. The rover seamlessly integrates advanced sensing systems with edge-deployed deep learning models for real-time weed detection and plant disease classification. Leveraging lightweight convolutional neural networks such as YOLOv5 Nano and MobileNetV2, Agrova achieves high-performance detection directly in the field, enabling precise interventions and minimizing manual labor. Field experiments demonstrate Agrova’s robust capabilities, with weed detection accuracy reaching approximately 89% and disease classification accuracy around 95%.