Predictive Quality AI in quality management: Use cases for an intelligent production environment

Related Vendors

Fabasoft Deutschland GmbH

FAULHABER Drive Systems

Taiwan Machine Tool & Accessory Builders' Association (TMBA)

Artificial intelligence opens many new doors in quality management. If the data basis is correct, the possibilities for application are almost unlimited. In this way, manufacturing companies come one crucial step closer to the goal of Predictive Quality.

* Andreas Dangl is an entrepreneur and managing director of Fabasoft Approve.

Applications are only as good as their data foundation. The importance of this IT wisdom, which also applies specifically to AI, is further underscored by the following paradox: Although the amount of data globally increases year after year, proportionally less information is available organization-wide, according to a recent Capgemini study. The main reason is the historically grown data silos within companies. Additionally, there is the problem that these contents are not consistently available across the entire supply chain.

An intelligent cloud-based document management system solves these challenges relatively easily and quickly. It stores all relevant information on a shared and comprehensively secured platform. Another advantage is that all players in the supply chain can be seamlessly integrated.

Once manufacturing companies have completed their homework regarding data availabilityQuality management through artificial intelligence: opportunities and benefits—ideally across all departmental and company boundaries—nothing stands in the way of moving towards quality management with AI support.

Numerous intelligent use cases

Artificial intelligence and its subgroups, machine learning and deep learning, offer a variety of applications in industrial environments.

The use of AI can start as early as the product development stage, where comprehensive knowledge of market and demand trends creates crucial competitive advantages. Automated analysis of relevant datasets by an intelligent system helps companies identify trends and accordingly adjust product development. Additionally, AI assists in the creation of new designs and concepts for machinery and equipment that not only meet defined quality standards but also offer innovations.

Another example is predictive maintenance, one of the most tangible applications in the realm of Industry 4.0. This approach allows for the extraction of condition data from machines and proactive maintenance of systems. The benefits are clear: businesses can improve their economics as downtime significantly decreases. Additionally, managers can define optimal maintenance times. Thanks to the continuous analysis of collected information, companies have the opportunity to sustainably improve the performance of their machines, thereby achieving higher productivity.

This is precisely where the biggest challenge of predictive maintenance lies: processing the vast amounts of data. This is where machine learning comes into play. By detecting the smallest anomalies, it is possible to predict failures before they occur.

Predictive Quality

In quality assurance, AI uses special pattern recognition to quickly and reliably identify defects. If the system detects a defect in the raw material, it can predict this in the final product, which in turn helps in sustainably avoiding such errors. Additionally, AI supports in specifying the defect.

With data-based prediction of product and process-related quality in manufacturing, companies are increasingly approaching the goal of predictive quality, which plays an increasingly important role in the constant tension between customer satisfaction and cost control. This can prevent defective products from entering the market, which is often associated with costly recalls and massive loss of reputation.

Improvement of supplier documentation

An intelligent document management system with AI support, as described above, serves, among other things, to connect partners along the supply chain. The players in the supply chain exchange order, quality, and project documents through digital processing, checking, and approval processes, and coordinate them with each other.

AI ensures that the review of documents provided by suppliers is more efficient than with traditional—and often manual—methods.

The smart software can validate test protocols and, if necessary, provide direct feedback to the supplier if, for example, they have referenced an incorrect part. It also automatically extracts the metadata contained in the documents and links them correctly.

Natural language communication

Modern plant documentation typically comprises thousands of documents. Managing and dealing daily with these mountains of data poses a challenge to users.

A modern document management system (short: DMS) offers intelligent interaction options for this purpose, based on Natural Language Processing (short: NLP). This allows technicians to intuitively communicate with the software during service operations on a plant and formulate technical questions about issues in natural language. The AI responds in the same way.

An intelligent DMS allows free choice of the underlying Large Language Models (short: LLMs). Whether it's ChatGPT from Open AI, Llama 2 from Meta, or models from Hugging Face, the AI generates answers from the company's internal data that are hardly distinguishable from human feedback.

AI-supported 8D process

An 8D report is a document used in quality management and exchanged between supplier and customer or internally in the event of a complaint. The designation 8D stands for the eight process steps required to address a defect to get to the root of the triggering problem.

The formal processes surrounding 8D reports require a lot of expertise from quality managers. AI provides essential decision-making bases in the various phases, thereby optimizing the overall process. This means that the intelligent system enhances the efficiency, accuracy, and speed of the business processes, thus increasing the quality of products and procedures while simultaneously reducing costs.

Another strength lies in the ability to take immediate actions. AI can be used to initiate automated processes for troubleshooting or to provide recommendations for improvement measures. However, the final decision on the actions to be taken still rests with humans.

Additionally, the AI is capable of deriving lessons learned from the immediate and preventive measures taken, which are available for addressing future, similar challenges.

Higher quality thanks to AI

Even the few examples of AI use in the production environment demonstrate the significance of this smart technology in quality management and related areas, and its role is set to increase even further in the future. The applications range from automating essential processes to intelligently integrating customers and suppliers into a common data environment.

At the same time, it is clearer than ever that the quality of the information base determines the extent to which companies can benefit from AI. This is precisely where a modern document management system comes in, helping not only to manage the growing daily data mountains but also, thanks to AI functionalities, to increase efficiency, productivity, and ultimately higher quality.

This article originally appeared on our partner portal www.industy-of-things.de (German Language), Vogel Communciations Group