Automated Quality Assurance System through E-TAMU for Hair Care Products

Most of the industries all over the world are using cutting edge technologies to save labor cost as well as time in many domains such as production, shipment, quality control etc. However, in Pakistan, majority of these tasks are done manually and are not time & cost effective. To overcome this

Project Title

Automated Quality Assurance System through E-TAMU for Hair Care Products

Project Area of Specialization

Artificial Intelligence

Project Summary

Most of the industries all over the world are using cutting edge technologies to save labor cost as well as time in many domains such as production, shipment, quality control etc. However, in Pakistan, majority of these tasks are done manually and are not time & cost effective. To overcome this hurdle, we propose a system, which will automate the task of quality control, which is currently being done manually. Automated inspection systems for quality assurance of different products are being used in almost every industry from automotive and electronics to food, pharma, cosmetics and many more. The replacement of manual inspection by AVI systems enable Industries to accelerate the process of quality check along with reduced manpower and reliable accurate results. Our targeted domain is to automate the manual inspection used in industries for quality control which not only has low accuracy rate but takes more time as well. Automated inspection process is guaranteed to reduce the error possibility and increase the inspection capacity in minimum time. The haircare product bottles of P&G are chosen for this purpose and an automated vision based application is implemented to detect the defects placed in the bottle and classify them according to the industry’s given inspection standard, ‘TAMU’ (Targeted, Acceptable, Marginal, Unacceptable). The minimum criterion for passing the inspection test is Marginal, and on a single unacceptable result, the product will be then discarded as faulty product.

Project Objectives

• To elevate the standards of quality assurance by promoting development of automated technological solutions to enhance testing capability and increase production quality.
• To acquire large amount of data for image processing and applying AI and ML techniques in order to increase the accuracy rate of inspection.
• To build hardware base system comprised of cameras and others controlling circuitry for Inspection along with the interactive GUI to display results

Project Implementation Method

• First step is for image acquisition in which a stable environment would be developed to provide static conditions. The images of defective and non-defective bottle will be acquired for processing.
• Next step would be to define region of interests for each of the specified defects for detection.
• Then the acquired data will go through certain pre-processing techniques. So that, the noise could be removed and designated areas for respective defect could be enhanced for better accuracy.
• Then the algorithms classical method or deep learning (if required) would be adopted to accomplish the goal.
• Then the algorithm will decide whether the provided bottle is defective or marginally defective or non-defective.
• The samples would be accepted and rejected on the basis of results.

Benefits of the Project

• Hardware aspect:

On the testbed, a high resolution camera mounted on a motor will capture images of the target product from different angles. There will have proper illumination around the product for better image processing. A controller circuitry will be incorporated to control camera position (motor) and illumination. The captured images will then be sent for defect analysis.

• Software aspect:

Images acquired from the testbed will be processed using classical image processing techniques as well as AI algorithms for optimized and accurate detection of target defects through which the product (bottle) will be classified as Acceptable (no defect found), Marginal (defect is negligible) and Unacceptable (defected is significant). These results will be visualized on an interactive Dashboard

Our target defects includes detection of any dent or scratch and presence of dirt on the bottle surface which may occur during manufacturing; bottle label alignment i.e. whether the label is placed correctly or not; to detect printing defects on label and bottle surface; bottle cap fault like whether the cap is seated properly.

Technical Details of Final Deliverable

• Hardware aspect:

On the testbed, a high resolution camera mounted on a motor will capture images of the target product from different angles. There will have proper illumination around the product for better image processing. A controller circuitry will be incorporated to control camera position (motor) and illumination. The captured images will then be sent for defect analysis.

• Software aspect:

Images acquired from the testbed will be processed using classical image processing techniques as well as AI algorithms for optimized and accurate detection of target defects through which the product (bottle) will be classified as Acceptable (no defect found), Marginal (defect is negligible) and Unacceptable (defected is significant). These results will be visualized on an interactive Dashboard

Our target defects includes detection of any dent or scratch and presence of dirt on the bottle surface which may occur during manufacturing; bottle label alignment i.e. whether the label is placed correctly or not; to detect printing defects on label and bottle surface; bottle cap fault like whether the cap is seated properly.

Final Deliverable of the Project

HW/SW integrated system

Core Industry

Manufacturing

Other Industries

Core Technology

Artificial Intelligence(AI)

Other Technologies

Sustainable Development Goals

Industry, Innovation and Infrastructure

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com