In this article, we will discuss data-driven quality control. Factory owners can utilize this article as a reference, in order to learn about the influence of quality control on customer satisfaction and the benefits of data-driven quality control.
This article covers:
Factory owners have always employed quality control procedures as part of their standard operations. This is vital in order to ensure that:
Factories that constantly produce and have a reputation for high-quality products all but guarantee brand loyalty among their customer base. This has a direct influence on increasing sales and the company’s bottom line. Furthermore, establishing a good reputation also assists with turning new customers into repeat buyers who place orders regularly with the factory.
Traditionally, quality inspections are conducted at the end of a production cycle, or alternatively, after defects are spotted during the production cycle. This results in additional costs since defects are only addressed after they actually occur and it can be difficult to ascertain at this stage which workstation or process in the assembly line was responsible for the defect.
Another key disadvantage of the standard methods of quality control utilized in factories is illustrated in the following example. Let’s consider a scenario in which a clothing manufacturer is producing a certain garment. The quality assurance specifications state that each garment should have a thread count of at least 220 and that all buttons on each of these garments should be subjected to 30 fatigue tests in order to ensure that they will not come loose. Additionally, each garment should undergo a seam strength test and be stretched along the seams and the edges to ensure that there is no separation of the seams.
If the factory at hand is producing a high output, to conduct individual quality tests on each of the garments produced by the assembly line will be a time-consuming and laborious process. These quality inspections can also lead to interruptions in other factory processes and cause extended delays. Also, if there is constant variation in a certain parameter – some garments have a thread count of 300, others have a thread count of 220 – this may be due to a certain machine needing maintenance, which the quality inspection will not pick up. This variation could be attributed to, or an indicator of fluctuations in machine performance; nevertheless all these garments will pass the quality inspection but the machine issues will not be addressed.
Furthermore, many of the traditional quality control inspections are carried out manually using checklists, and the results of the inspections are only entered into data management systems after all the quality inspections are completed.
Data-driven quality control refers to:
Data-driven quality control is an alternative option to traditional quality inspections which involve conducting large numbers of individualized tests on each product, post-production. There are slight impediments to data-driven quality control in the case of brownfield sites, where various protocols are employed and there are still old contracts of sale with machine manufacturers. Additionally, when factory owners are considering implementing data-driven quality control, they should bear in mind that there is a need to have a robust, industrial cloud that can bring in all this data.
The benefits of data-driven quality control are numerous. Firstly there is the reduction in the time since samples are compared in an automated manner to a model, so the standard delays are reduced. Another key benefit is that the entire product lifecycle can be monitored and analyzed continuously, so potential defects can be detected before they happen and/or corrected early on, rather than just at the end of the lifecycle.
The third major benefit is that quality control processes are not operating in isolation. Instead, they are integrated into the main factory operations. For example, in instances where a certain machine’s performance is causing variation in products, machine learning models from the predictive maintenance programs in the factory can be integrated into the quality models. Consequently, the factory personnel can make truly informed decisions and receive multiple insights from the data-driven quality control processes.
Data-driven quality control allows for the real-time integration of multiple sources of external quality-related data. A factory owner could integrate real-time customer responses on social media to a certain product, or reports about defects customers are encountering into the quality models. This allows customers to become part of the factory quality processes and factory owners to speedily address post-production issues. Customers will then feel that their inputs and feedback are being addressed and this will have a net positive effect on the overall customer experience.
A smart factory or manufacturing environment refers to a situation in which:
Many factory owners are trying to transform their factories into smart factories since this means increased efficiency, productivity, the reduction of waste in their factories and the identification of new sources of revenue.
It can be tempting for a factory owner when considering implementing big data solutions to focus solely on increased production while neglecting or missing other areas that data can also enhance. Quality control has in the past operated in isolation and not benefitted as much from data solutions. However, in order to increase revenue – production, customer satisfaction and sales should increase. Effective data-driven quality control can contribute in a major way to an increase in customer satisfaction and sales, so it is worthwhile to not neglect its integration into the factory platforms.