Intelligent Product Supply System Using Artificial Intelligence with YoloV5

This paper explores the integration of wireless sensor networks (WSN) and smart city solutions in the realm of Internet of Things (IoT), specifically focusing on enhancing consumer-facing systems. The study introduces the implementation of the YOLOv5 supervised learning-based object detection model to conduct image inference for product detection on supermarket shelves. Utilizing the ESP32CAM module for image capture, the proposed approach automates the product re-supply process, offering insights into consumer behavior. The system employs a wireless communication interface developed with Wi-Fi, Raspberry Pi, and Microsoft Azure to relay information about product availability. Results indicate the model’s efficacy in object detection, boasting a mean average precision (mAP) exceeding 0.5 and achieving 96% accuracy in determining product availability. The research marks a substantial stride in supermarket stock management, utilizing IoT and supervised learning to enhance customer satisfaction and optimize sourcing processes.