Condition Monitoring in the Cloud
Challenges
The technical availability of production systems is playing an increasingly important role due to high quantities and tightly calculated profit thresholds. Particularly high economic risks and damage result from unplanned machine downtimes.
The failure frequency is not a constant, but a parameter to be minimized by maintenance strategies. Condition monitoring and the associated condition-oriented maintenance represent a valuable approach to reducing machine downtimes.
From the point of view of machine manufacturers, however, the application of condition monitoring currently poses major challenges:
1) Meaningful wear models with temporal prognostic capability must be developed.
2) Differentiated wear models must be combined with machine and process models so that wear-adaptive process parameterization becomes possible.
Approach
New networking technologies such as the cloud offer great potential. As a central platform, it enables optimized modeling based on an extensive database in combination with powerful algorithms as well as access to extended component, machine and process models. This allows new services to be offered, which in turn extend the local machine functionality.
Research within the Cluster of Excellence focuses on the conception and realization of a reference architecture. Both the local connection of machines with dynamic information models (OPC UA), communication aspects and in particular the cloud itself are considered.
As a result, e.g. production-technical apps can be subscribed for a state estimation or the calculation of optimal process parameters.