DESIGN AND IMPLEMENTATION SERVER SUB-SYSTEM AND PROCESSING AND DATA VISUALIZATION FOR AUTOMATIC VISUAL INSPECTION SYSTEM IN PACKAGIN CARTONING MACHINE

• Main Author: Noorsyach Achva, Raka
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/70582
• Institution: Institut Teknologi Bandung

An automatic visual inspection system is a system that is being developed in cosmetic manufacturing companies, especially in product packaging (cartoning) lines. This system is important for companies to maintain product quality standards as the speed of the product packaging process increases. The automatic visual inspection system will detect product packaging discrepancies in the packaging process, and then the system will remove the product from the packaging line. Stakeholders in the company need to know the quantity and time when mismatched product packaging is found because the data is valuable for the company to analyze which company can obtain important information resulting in product packaging process improvement. Therefore, it is necessary to have a system to visually present the inspection results from the visual inspection system as a benchmark for the performance of the packaging process occurring in one work shift. The server sub- system is one of the automated visual inspection system components that stores inspection results data in a database and deployment of web application. The method used in designing this system is capstone design, which is a system design method based on the needs and problems of prospective users. In particular, users expect a platform, in this case, a condition monitoring platform, to monitor the inspection process running on existing devices owned by the company, with a 1920 x 1080 pixels resolution. User satisfaction with the interface display must exceed the standards determined globally using the system usability scale (SUS) method with a minimum score of 67 points. The realization of the server sub-system lies in two things. Implement a new database to store inspection and interface display for inspection results. The database stored in this sub-system is integrated with the company's master database. This new database stores information in the form of inspection results carried out by the processing sub-system. Database implementation is done using PostgreSQL as RDBMS and followed by SQL-based data manipulation platform, PgAdmin4. For interface display, a PHP-based back-end framework is used to fetch data and communicate web applications with databases and processing sub- systems using the HTTP communication protocol. Web application development uses a PHP-based framework, namely Yii2. Yii2 as a framework adopts MVC, or called as model-view-controller, as its software architecture development design. MVC consists of 3 components. The model has ability to store data, manipulate data from databases, and authenticate users. View has ability to store business logic that is performed on server-side rendering, generally this business logic is written using JavaScript and the execution of business logic is carried out without the need to refresh the page. Lastly, the controller has ability to act as a storage API service and generally performs integration between requests on the client-side/view component and responses from the database on the component model. The selection of the PostgreSQL database was carried out by considering the ease of scalability and maintenance considering the huge support of PostgreSQL community as well as the easy integration between the company's existing master database and the automated visual inspection system database. The selection of a web application development framework using Yii2 is also based on the same considerations because the applications supporting the enterprise manufacturing are based on the Yii2 framework. Testing the implementation of the web application consists of measuring the screen resolution of the user's device, as well as the results of a survey on the use of the interface display on the web application using the system usability scale (SUS) method. The measured layer resolution at implementation is 1920 x 1080 pixels or FHD, while the minimum specifications are 800 x 600 pixels. Also, the results of the survey on the ease of use of the interface display meet the specifications because the value of the survey results is 78 points while the minimum specification value is 67 points. In general, this project has provided a broader understanding and utilization of web applications, because the use of web applications is not only limited to commercial use, but can also be implemented in the company's manufacturing processes.