Chandeepa Vidusara Kumarasiri Verified Automation Engineer

Abstract: Early detection of agricultural crop diseases is critical for food security and yield optimization. This research developed an end-to-end autonomous UAV quadcopter framework equipped with calibrated RGB and modified near-infrared (NIR) sensors to capture multi-spectral canopy imagery over paddy fields. A tailored convolutional neural network (CNN) model achieved high diagnostic accuracy in identifying bacterial blight and brown spot diseases at early onset, integrated with a lightweight Flask web platform for real-time agronomic advisory.

Abstract: This paper presents the design, kinematic modeling, and implementation of a microcontroller-driven closed-loop cutting system tailored for high-volume automated industrial manufacturing. The architecture combines precision stepper microstepping, real-time optical pulse feedback, and custom firmware algorithms to ensure sub-millimetre cut repeatability while minimizing mechanical wear and production downtime.