Development of a Non Invasive RPW Management Detection Tool Using Machine Learning Approach

One of the main pillars of the economies of many nations in the world is palm planting. Palms are a significant part of a profitable agricultural industry in India as well. Any danger to the palm trees’ output will result in a catastrophic economic collapse. Red Palm Weevil (RPW) is considered the most important pest of the palm trees. The RPW attacks the young palm trees less than 15 years old and lays their eggs in the crown of the tree. The larva that emerges out of these eggs lives inside the trunk of the palm tree by feeding on their soft tissues making it difficult to identify. The aim of this project is to develop a non-invasive acoustical instrument that could sense the acoustics produced by the larval movements. In this research, the instrument is trained with different acoustics of the larval movement making it possible to detect the presence of larva by differentiating it from the background noises. The software was designed to record and analyze the sound in the real-time environment and it was successfully tested in various fields. The device was capable of detecting the larva at an early stage and showed an efficiency of 98%.