Publication

Abstract : The growing universal population and rising demand for food are putting substantial pressure on the global agricultural system to boost production. However, expanding farmland often comes at the expense of wildlife habitats, escalating human-wildlife conflicts as crops are damaged by animals, threatening food security. Furthermore, the use of unsustainable deterrent methods can harm biodiversity. Current crop protection methods fall short of supporting sustainable agriculture goals. This study focuses on mitigating human-wildlife conflict by introducing an IoT-based system that integrates AI-driven deep learning for real-time farm management in India. The proposed system employs an Ultrasonic Sensor A01NYUB to detect motion near farms, triggering a camera equipped with a custom-trained YOLOv8 deep-learning model that identifies animal species with 99 % accuracy. Based on the classification, species-specific deterrent actuators safely repel animals. The innovation also incorporates “recovery zones,” designated areas with food and water to redirect animals away from farms. These zones are monitored by an HC-SR04 ultrasonic sensor that is used as a simple counter-interfaced with an ATtiny85 microcontroller to track animal entries. The system demonstrated its effectiveness by reducing crop damage and fostering coexistence between humans and wildlife. Combining IoT technology with indigenous knowledge provides real-time monitoring and enhances both agricultural productivity and biodiversity conservation. This scalable and adaptable solution contributes to Sustainable Development Goals (SDGs) 2 (Zero Hunger) and 15 (Life on Land), representing a key innovation in precision agriculture and sustainable farm management.