(QAOD) Quality Assurance using Object Detection – ...
Enterprise Resource Planning Blogs by Members
Gain new perspectives and knowledge about enterprise resource planning in blog posts from community members. Share your own comments and ERP insights today!
No matter what the size of the industry, it is no secret that technology has made a profound impact on business operations in terms of productivity and efficiency. New advancements are strategically positioned to revolutionize the economy and develop the world to become market pioneers of the industry.
A perfect example of an industry that relies on well-structured systems is the textile industry. It is one of the largest sectors that employ millions of people all over the world. The Philippines plans to bring a resurgence of a reputation as a competitive player in both domestic and international markets.
Madison (2019) states that the "textile industry is a global industry which is one of the largest in the world and employs millions of people. Textile materials are made from fiber, yarns, and fabrics. The ability of a country to create new machines and technology gives its manufacturers an even larger share of this market, which, in turn, results in greater profits."
The textile industry used to be the top-performing sectors locally and internationally, according to Rodolfo (2018), the Trade Undersecretary for Industry Development and Trade Policy Managing Head, with its strategic initiatives, projects, and organizations, including key players, makers, and partners in progress. "Garments and Textile Export Board" (2019) agrees that the Philippine textile and garments industry is one of the country's success stories, making the quality of the textile highly significant.
Maintaining an adequate standard of quality also costs effort, and that is the end goal of any production and manufacturing processes; to have a high quality and reliable product as results. It is crucial to understand that it is a waste of time and resources if the quality of the outcome does not comply with the expectations and needs of the customers.
Quality Assurance using Object Detection Solution Demo:
Challenge Submission
Solution Name:
Quality Assurance using Object Detection and Arduino Uno
Solution Description:
The main reason why we decided to develop a Quality Assurance System using Object Detection and Arduino Uno that will contribute to the textile industry in assuring the quality in terms of detecting dirt stain, oil stamp, spot stain, fly yarn, shaded, loose thread, dye mark, hole, bird's eye, slub, red stain, drop needle, and needle run. It will also help the textile industry to generate a report regarding the detected image. It will then automatically stop the machine when the system detects a suspicious object.
At the back-end, the data is integrated into SAP Business One. Once the system detects an object, it sends an activity to the Item Master Data and Business Partner Master Data on SAP. The system generates a report about the number of detected suspicious objects, a list of detected images per item, and a list of suppliers who detected a malicious object.
Solution Use Case
Real Customer Use-case:
The fabric inspector do the following:
Select item that will be undergo on inspection
Monitor the system if their has a detected image
View the generated report
Persona Identified:
Pain Points:
Manual checking of holes, stain, and damage in textile
Manual Recording Of Damage Textile
Industry Focus:
Textile Industry
Solution Details:
Quality Assurance using Object Detection and Arduino Uno is a web based application which helps any manufacturing for textile and fabric industry
Object Detection features intended to automate the process of checking the textile if it has a stain, holes, or any damage in the textile.
Solution Technology
SAP Cloud Platform
Machine Learning and Internet of Things
C#
MVC Framework
Tensorflow
Road Map
Long Terms Plan :
Implementation of robotics in tagging of sticker when the system detected a suspicious image.
Send an alert via email/sms regarding generated report of detected suspicious object daily, weekly, monthly, quarterly,yearly(it depends on the setup of the system admin)
Add predictive analysis upon QA, the system will suggest a pattern that will be plotted on the textile based on the target quota.