Möchtest du unsere Inhalte auf Deutsch sehen?


Monitor manufacturing in real time and reduce scrap

IoT Use Case Schaeffler
4 minutes Reading time
4 minutes Reading time
In order to save costly and time-consuming quality assurance measures in manufacturing processes, the workpiece quality must already be monitored during ongoing production so that defects are not only detected at the end of the process chain. Interventions in the process should ideally take place directly in real time. Schaeffler Digital Solutions uses the example of transmission production to explain how manufacturing errors can be avoided through vibration analysis in real time.

The challenge:

Fine machining processes such as honing or grinding are used in transmission manufacturing. These methods can cause problems, such as periodic geometry deviations like microwaviness. This is caused by vibrations during the manufacturing process due to relative movements between the tool and the workpiece. Microwaviness often lead to noise problems with gears. If these are not detected until the end-of-line test, the entire transmission would have to be dismantled again in case of doubt. Another problem with honing is honing ring breakage. If this event is not noticed in time, fragments of the broken honing ring can cause consequential damage in the machine. In addition, defective parts would remain in the production process. In the practice of machining, user errors are also possible. Clamping errors can occur during the dressing process – imbalances when the spindle starts up are the result.

The solution

The goal is to control parts in the machine across all work steps and manufacturing processes, in real time during production. Software solutions such as autinityVC from Schaeffler Digital Solutions do just that: autinityVC is a vibration analysis software that evaluates data from vibration sensors during ongoing production, compares it with stored reference values, and reacts immediately in the event of deviations by accessing the machine controller via an interface. For example, a tool breakage would cause a sudden peak in vibration, so autinityVC’s shock monitoring system would sound an alarm, trigger a notification, and stop the machine. The software also automatically detects components that are to be considered acoustically critical due to pronounced microwaviness. The affected workpiece is rejected or marked for tracking.

The software is installed on an industrial PC directly at the machine. It records the analog signals of several vibration sensors, decomposes them into individual frequencies using the Fast Fourier Transform (FFT) and compares them with previously stored reference spectra. If deviations from limit values are detected on the basis of these target/actual comparisons, an alarm is triggered. In addition, autinityVC records required process parameters from the programmable logic controller, which are automatically linked to the acquired spectra. Thus, work-step and material-specific references can be created and compared. Direct communication with the machine allows immediate ejection of parts and emergency retraction in the event of a shock alarm. All spectra and shock data including OK/NOK evaluation are stored locally or on a network drive. For further data analysis and visualization, the “VibroAnalyzer”, a desktop tool with Campbell diagram, trend evaluation and filter functions, is used to detect anomalies in the vibration image. Third party systems can also be used for analysis as all vibration data is available in CSV and/or Q-DAS ASCII transfer format. For central data storage, notification function and display of NOK trends, a server solution is used that combines all connected machines at a glance. The recorded data is made available on-premises – this means that the sensitive data is stored locally and not on external servers. This ensures a high level of data security. If required, Schaeffler Digital Solutions also provides off-premises solutions (cloud/data center).

The result

The quality of parts and material is determined in real time during production. Defective parts can be rejected without passing through further production steps. Using the example of process monitoring on honing machines, considerable repair costs per damage case can be saved by using autinityVC. In the event of any production errors or tool breakage, autinityVC stops the machine and initiates the ejection of the affected parts. The subsequent workpieces remain intact – scrap is reduced. The rapid detection of deviations in the production process prevents consequential damage and thus long, unplanned downtimes. autinityVC offers an analysis tool that can be used to evaluate and improve production processes – thereby optimizing cycle time. Furthermore, tool wear is detected and the tool can be replaced in time. autinityVC can also be used as a data source for traceability, as it allows parts to be identified downstream. 

In application

Get our IoT Use Case Update now

Get exclusive monthly insights into our use cases, activities and news from the network - Register now for free.