Artificial Intelligence and Internet of Things (Ai-IOT) Enabled Mobilenet Model for Climate-Resilient Smart Livestock Farming
Abstract
In a world where rapid advances in technology are obviously shaping all spheres of life, the agriculture and livestock management field is not exempt. The integration of AI and IoT-based technologies has birthed a new world of smart systems capable of revolutionising the agricultural industry for a more efficient and sustainable farming system. This research considered the AI-IoT-based model for advanced monitoring of animal (sheep) health with the objectives of reducing latency of the existing systems, increasing productivity, enhancing disease management and ultimately improving livestock welfare. This was achieved with the help of the TeSLF Embedded system, which practically contains sensors such as those for weight, humidity and temperature (for the animals and the environment, independently) and for the purpose of data collection. These data were gathered in real-time from the farm and transmitted through an edge gateway to reduce the latency of the existing system. They were then analysed by the Arduino Nano BLE Sense microcontroller to generate a compatible format for the Google Teachable Machine responsible for training the system using the MobileNet. The trained model was then sent to the ThinSpeak platform for farmers, caregivers, veterinarians and researchers to monitor the livestock’s activities remotely and in real-time on their mobile and handheld device TeSLF app. The TeSLF embedded system offers automated data collection, analysis, and real-time monitoring for timely disease detection and intervention, leading to improved health of the entire herd. The model was evaluated using accuracy, precision, and F1-score. It produced an accuracy of 98% against that of the existing system, which produced an accuracy of 88%, deployed on edge devices for on-site decision-making, enabling farmers to mitigate heat stress and optimise animal welfare under varying climatic conditions.