KI Use Cases in der Produktion: Lebensdaueroptimierung von Produktionsmitteln
How can AI help extend the lifespan of production equipment? This is the question we explore in Part 2 of our mini-series on AI use cases in manufacturing. Each of the presented AI use cases builds on the 4 opportunities for value creation with AI in industry.
Every manufacturing process requires a range of production equipment. Some production equipment is highly complex and expensive, such as tools like spindles, rollers, or casting molds. With increasing usage duration, this equipment wears down. Depending on process management and load intensity, wear occurs at different rates.
Beyond a certain degree of wear, quality problems can occur in manufactured parts, or even cracks and deformations in the tools, and in worst cases, damage to the machine itself.
AI can help identify increased equipment wear early and slow it down by adjusting process management. In critical cases, AI can trigger a stop to the manufacturing process or request maintenance or tool replacement.
Sensors and a monitoring system pave the way for AI
The starting point for AI development is an AI-ready data inventory. In a manufacturing context, this data inventory is typically generated through one or more monitoring systems, which capture production equipment and system states in real time and make the data accessible for further analysis. For equipment lifespan optimization, it is often necessary to first equip the manufacturing system or production equipment with additional sensors to generate data from which conclusions about the condition of production equipment can be drawn.
Stage 1: AI Detection
In the generated dataset, a detection AI can search for patterns in the progression of wear processes and analyze these in relation to process data and process settings. The results can be provided to machine operators in the form of an alert system that activates whenever the current wear pattern reaches a critical range.
Stage 2: AI Diagnosis
In the next step, a diagnostic AI uncovers causal relationships by linking wear patterns to the underlying process data and process settings, particularly target-actual deviations. In diagnosis, product quality criteria can also be considered if they are available in real time.
As a distinguishing feature compared to conventional approaches, the AI learns the rules of the underlying causal chains purely on a data basis, meaning it never receives an explicitly defined rule.
The way in which an AI stores and outputs implicitly learned rules is often incomprehensible to humans, as simple "if X, then Y" relationships rarely emerge. AI diagnoses can therefore be quite unsatisfying for humans. aiXbrain therefore works intensively on human-centered AI that provides, among other things, explanations understandable to humans.
Stage 3: AI Prediction
A predictive AI now reverses the rules learned in AI diagnosis in time and converts observed process data, process settings, and the current production equipment condition into predicted wear progressions. This makes it possible to recognize in advance when it will become critical in the future, based on which current wear patterns and process settings, for the condition of equipment, systems, and potentially product quality.
The results can be used in an information system for machine operators and production planners. For each production equipment, there is a forecast of the expected remaining lifespan and a list of features that the AI uses for its assessment. When things become critical, the AI can proactively request maintenance or tool replacement, or even recommend stopping the manufacturing process.
Stage 4: AI Control
In the final expansion stage, a control AI actively makes suggestions for optimizing the lifespan of production equipment. Once an AI has seen enough setting-wear pairs, it can independently draw conclusions about which adjustments to process settings lead to which changes in wear patterns. If the predictive AI forecasts that the current wear progression with unchanged process settings will lead to shortened lifespan, the control AI can submit a setting suggestion for the process that breaks the current wear pattern or proactively slows wear.
Value of AI for Equipment Lifespan Optimization
Keeping track of the lifespans of all production equipment and, beyond that, optimizing their lifespan is a task that cannot be accomplished manually. Avoiding accelerated wear is not possible without comprehensive understanding of causal chains.
AI in its various forms of application shines light into corners that humans currently cannot see, and thus enables sustainable improvements in process management.
For daily operation use, AI technology can be provided in the form of one or more AI assistants. This gives machine operators a powerful tool that enables them to work together with AI to master complex challenges in process management. aiXbrain places great value on establishing such hybrid intelligence systems, in which humans and AI collaboratively complement each other.
If you are interested in learning about further value-creating AI use cases in manufacturing, we recommend our other articles in this mini-series:
