A Deep Learning Based Automated Defect Detection System for Manufacturing Industry
This project is conceived on the same model and the domain of the deep learning based automated system is used as an for detection of defect lead acid plats. The project defects inspection is an active problem of Lead Acid Battery manufacturing in the associated industries.
Salient Features
The Deep Learning-based Defect Detection System (DLbDDS) is a generalized solution for defect detection manufacturing items in industries. As a proof of concept, a typical Lead Acid Battery (LAB) is considered in this funded project in a manufacturing environment to detect defective battery plates. The salient features of the project are:
IoT-enabled lead acid plate-specific conveyor belts
A deep learning for model training and real-time plate recognition that distinguishes between rotting and good plates
An artificial hand that uses the model decision-making to distinguish between the good plates and the bad ones
The data analytics and prediction are integrated into a smart dashboard.
Controlling the conveyor belt’s speed and hand movement with an HDMI panel.
Each decision about the good and bad plate is recorded, including the time, status, and batch on the cloud that is accessible from anywhere at any time
Reporting and alerts when specified problem levels are reached
Scalability of Project
The project can be scaled and modified according to the requirement of manufacturing industries. The IoT-enabled conveyor belts can be designed to specific product requirements. With the same deep learning algorithm, the model can also be trained for new products.