How to Achieve Zero-Defect Manufacturing with Bodhee® Predictive Quality AI App

Bodhee® Predictive Quality AI App for discrete manufacturing optimization predicts faulty outcomes ahead of time and eliminates the scope of rejection of goods. Read how you can achieve Zero Defect Manufacturing and a real-world use case about how business benefits were delivered.

Work SMART, not hard; you need to correct your production process, not your people!

Discrete manufacturing is a complex environment where products get assembled on a production line in small batches, usually customized to specifications, and even a single defective product is unacceptable. When the production process gets complex, there is greater scope for errors leading to defects in a product. The slightest defect can cost the company precious time and money as the product gets scrapped. Quality production is essential in discrete manufacturing, as rejected defective goods can only mean lost profits and unhappy customers. Maintaining Zero Defects Quality (ZDQ) standards is crucial to keep up with the competition, too.

What is Zero-Defect Manufacturing?

Zero-defect manufacturing (ZDM) is a quality management philosophy, where all steps in the production process ensure that the resultant goods do not have any defects. Non-conformance to stipulated quality standards is recorded as a defect, and every process gets scrutinized to detect the cause. Besides the primary goal of improving product quality, zero-defect manufacturing can secure significant financial savings. Zero Defect Manufacturing aims to eliminate the losses incurred from the rejection of defective products and reduce labor hours wasted on inspection and rework, thus increasing productivity.

Zero-defect manufacturing has its roots in an idea sparked by Dr. Genichi Taguchi in the 1960s that turned Japanese manufacturing on its head. However, it was not until the 1980s that the United States and Europe began to see how their ways to ensure quality manufacturing were redundant compared to the Japanese. The old methods relied heavily on inspecting products as they rolled off the production line and discarding those pieces that did not meet a certain standard. However, Dr. Taguchi astutely pointed out that no amount of inspection could improve a product; quality had to be designed into the product from the start. [1]

ZDM is a holistic approach for ensuring process and product quality by reducing defects through corrective, preventive, and predictive techniques. It mainly uses data-driven technologies, guaranteeing no defective products leave the production site and reach the customer, aiming at higher manufacturing sustainability.

-This definition is a result of a CEN-CENELEC ZDM terminology standardization process. [2]

Championing this thought and taking a step further in this age of Industry 4.0 and the Industrial Internet of Things, Neewee is committed to providing our clients with Data Analytics and AI apps for manufacturing optimization. 

Our Bodhee® Predictive Quality AI App is an innovative technology designed to become an integral part of the entire discrete manufacturing process. The solution to ensuring zero-defect quality manufacturing is now most easily accessible. 

Bodhee® Predictive Quality AI App: Significant Commercial Benefits with Zero Defect Manufacturing

Zero defects! That may sound like an unattainable feat for many organizations, but not for manufacturers using our Bodhee® Predictive Quality (AI) App. ZDM is now easier to implement by leveraging big data algorithms and machine learning capabilities. Our advanced data-driven AI app has an extremely user-friendly interface that does not require special skills and expertise, saving a significant amount of time on implementation. 

Due to the uncertainty and variability threatening production, it is a common argument that ZDM is not feasible. For example, natural raw materials often come with inherent variable complexities, which generates a defect that will impact and propagate throughout the production line. In such a case, ZDM involves considerable effort in tracking the quality of the product at every stage in production. It can get time-consuming, costly, and even result in wastage. Bodhee® Predictive Quality (AI) App also demolishes all these hurdles by achieving what is deemed virtually impossible. 

  • Improved Quality Control- Bodhee® Predictive Quality (AI) App helps break down data silos. Seamless integration with the complete manufacturing system and interoperability with other IIoT platforms paves the way for quality-oriented approaches and improved process control.
  • Prediction of Upstream Disturbances- With data-powered end-to-end visibility of the manufacturing process, the AI app can predict a disturbance or deviation in the workflow in the immediate future, which enables the production management team to prevent it before it occurs. Real-time insights forecasting product or process variation can both facilitate defect minimization at the end of the production line.
  • Increased Productivity- Bodhee® Predictive Quality (AI) App has a two-step process of macro and micro-interaction mapping. While performing micro-mapping, the AI app carefully charts out how individual parameters impact one another across workstations. During the macro mapping stage, the app determines how these micro-level changes affect the quality of the raw material, in-process products, and the final product. Thus, it ensures the yield is defect-free every time, improving the throughput, too.    
  • Reduction of Wastage- Backed by Machine Learning algorithms, the AI app ingests LIVE production data—product and bill-of-material details, work orders, production flow updates, etc. Then, our trained Risk Score Algorithm, specifically designed for discrete manufacturing, assigns a risk score at each downstream workstation as the product moves through the manufacturing process. If the risk score crosses a threshold during the process, the app immediately sends out alerts and recommends downstream corrective actions for each workstation and quality gate. Thus, quality managers need not wait for the final output to do a quality check for defects. Predictive assessment and indications can help reduce the scope for errors and prevent wastage of product, time, and labor.   
  • Cutting Cost – Timely remedial actions that prevent wastage of all resources mean cost savings. Also, all the production data gets fed back into the system. The Reinforcement Learning algorithms that are ALWAYS learning work their magic, optimizing the future production processes. Efficient processes and zero wastage correspondingly lower production costs.
  • Improved Work-order Closure Rate – Bodhee® Predictive Quality (AI) App facilitates resilient production through flexible processes that can quickly adapt to any changes that need to be incorporated. Prediction of possible errors can prevent a defect from creeping into the product. Eliminating the chances of rejection helps reduce reworking and saves time at the quality gates.

Designed keeping in mind High-Variation, Low-Volume Discrete Manufacturing, the AI app takes only 6-8 weeks to deploy impressive ROI for a medium complexity line.

Here is a use case that will give you a more detailed idea of how the app has helped achieve manufacturing goals in the real world.

User’s Story….

The customer is one of the largest cigarette manufacturing companies in the world. Their product manufacturing process comprises of two parts. In the first part, freshly harvested raw material gets cured and processed to obtain cigarette-grade tobacco. The second part of the process involves the rolling of cigarettes and packaging.

The production process in focus here is the first part where manufacturers need to treat and age the tobacco appropriately to enhance the flavors and aroma. At different stages of the process, moisture must be uniform and at pre-determined levels in the tobacco. The moisture levels significantly impact the overall quality and throughput yield, and therefore, it is a crucial parameter.

The Business Problem:

Since leaves from different types of tobacco get blended to create a particular flavor, and because the natural raw material is highly hygroscopic, it was getting challenging to maintain the correct moisture levels. There was a natural variation in the inherent moisture of tobacco due to the region it came from, season, atmospheric condition, the way the tobacco got stored, etc. The operator found it hard to determine the ideal steam pressure, adjustment of the water pressure, how many valves to open, etc., as these factors not only influenced the texture and qualities of tobacco but affected its grade, too. The output quality and quantity of tobacco were heavily dependent on moisture level post-primary conditioner.

The thrashing process produced a lot of waste since they failed to maintain moisture at appropriate levels. That also had a severe impact on the color of the tobacco, which in turn affected the quality.

Thus, the fluctuations in the moisture levels of tobacco had two consequences—1) Low Production and 2) Low Quality.

How Bodhee® Predictive Quality AI App Rolled Out the Solutions:

With a holistic view of the process, our AI App provided predictive insights and recommendations in real-time for course correction. Even though it was a fast-moving process, the AI app augmented process control. As the tobacco moved on the conveyor, the Reinforcement Learning algorithm collected learning about the moisture level and the resulting output quality at the end of the process. Based on the data inputs from the various processes parameters, our Predictive Quality AI app determined and recommended the ideal values to optimize process control, generating and maintaining appropriate moisture levels. Precision recommendations, such as the number of water and steam valves that must be opened, guided speedy process correction and helped achieve the desired results. Thus, significantly reducing the standard deviation of the moisture against the set points and preventing impact on the production costs.

With the help of our advanced algorithms and the Predictive Quality AI app, the production management team was able to identify and correct the errors in their production process. Our AI solutions helped the cigarette manufacturing company achieve zero-defect production, ensuring that every pack of its products was of the highest quality.

Business Benefits Delivered:

  • Optimized Productivity by 6 %
  • Improved Quality of Output by 18 %

Predicting the future isn’t magic, it is Artificial Intelligence! -Dave Waters

Manufacturers can use predictive analytics to take the trial-and-error out of their production campaigns. Our AI/ ML technology is revolutionizing the manufacturing industry with its ability to provide real-time data on production quality, improve efficiency, reduce costs, and deliver optimum yields. Additionally, adoption and implementation have proven cost-effective for SMEs and large enterprises.

Schedule a Demo for the Bodhee® Predictive Quality AI App to learn more now!

References:

[1] Simpson, T. W. (2000). 32.3 Taguchi’s Robust Design Method. https://www.mne.psu.edu/simpson/courses/ie466/ie466.robust.handout.PDF

[2] Psarommatis, F., Sousa, J., Mendonça, J. P., & Kiritsis, D. (2021). Zero-defect manufacturing the approach for higher manufacturing sustainability in the era of industry 4.0: a position paper. International Journal of Production Research, 1–19. https://doi.org/10.1080/00207543.2021.1987551

Related Blog

How 1 Amazing AI/ML Strategy Can Propel Aerospace Manufacturing

“Eureka!” yelled Archimedes and went down in history as the greatest mathematician of antiquity for making that one phenomenal discovery. Read more

Forrester Lists Neewee Amongst Digital Twin Service Providers in 2022 Report

Neewee is honored to have been named on a vendor list of Digital Twin Service Providers published in the prestigious Read more

Micro Scheduling AI: Unpacking the True Benefits of Scheduling for F&B Manufacturing

We believe that AI apps for timely packaging and delivery are the future of the F&B manufacturing industry. Read to Read more

How Does Industrial Edge Computing Increase Your Competitive Edge?

Is there a solution for the heavy computation and industrial big data analytics that the manufacturing sector requires? Centralized data Read more