An IOT Integrated SSD Model for Detection of Faulty Containers in Beverage Industry

In beverage industry, ensuring quality of cool-drink bottles traditionally involves a combination of sensors and manual inspection. This dual approach is prone to errors and decreases product quality and customer satisfaction. Our project focuses on improving this process through "IOT integrated single shot detector" (SSD) model for detection of faulty containers in beverage industry, aiming to enhance accuracy and efficiency in quality control of bottles. To address this challenge, we implement a "Camera-Based System" to replace traditional sensors with "Deep Learning Technology" in quality control of bottles which can offer several advantages. In this new system, we trained SSD model with a dataset to predict faults in bottles during production. The SSD model is chosen for "Speed and Real-time Use": One of the key advantages of SSD Model is its speed and generates spontaneous predictions. This makes it highly suitable for real-time applications. When a fault is detected, the model sends information to an "Arduino Microcontroller" which is responsible for controlling servo motors. Bottles without faults are packed, and bottles with faults are removed using a servo motor positioned along the conveyor belt for refilling. This ensures that only high-quality products move forward for distribution to consumers.