DEVELOPMENT AND IMPLEMENTATION OF AN AUTOMATIC MONKEY DETECTION AND REPELLENT SYSTEM TO ENHANCE AGRICULTURAL PRODUCTIVITY IN REMPANG CATE, BATAM

Abstract

Monkey pest attacks pose a serious threat to agricultural productivity, particularly affecting the Tunas Baru Farmers Group. This study aims to implement appropriate technology to mitigate crop damage caused by these pests. The activity was conducted in Rempang Cate Ward from May to September 2025. The implementation method included field observation, system design (hardware and software), on-site installation, as well as monitoring and performance evaluation. The developed device utilizes a Convolutional Neural Network (CNN) with a Raspberry Pi microcontroller and a webcam to detect the presence of monkeys within a radius of up to 35 meters. Upon detection, the system automatically generates a gunshot sound audible up to a radius of 250 meters to repel the pests. Given that the partner's location lacks access to the PLN electricity grid, the device is powered by solar panels as a renewable energy source. The implementation results demonstrate a significant impact, marked by a reduction in crop damage from an average of nine trees to only three trees per attack incident. This reduction contributed to a 67% increase in production yield compared to the pre-installation period. This technology has proven to be a more effective and efficient solution than manual guarding methods and has the potential to be replicated in other areas to support food security and rural welfare.