| Use Case | Description and Exploitation |
| Big Data Analysis and Mining | Big Data analysis methodologies enable e.g. the identification of associations and patterns in data (streams). This is useful to increase the knowledge about analysed manufacturing systems and processes, like relationships between happenings, influencing factors or causalities. It is also usable to extract trends potentially useful for humans and systems on several subjects. |
| Monitoring and Observation | Big Data analysis enables monitoring and observation based on defined patterns. This could provide hints about anomalies (system manipulations, hacking activities, energy leaks, etc.), failures, or other happenings. Pattern matching, through defined thresholds, could trigger notifications (e.g. if a pattern reaches a defined threshold then notify an operator or/and a system). Patterns could also represent a KPI represented on a dashboard. It is also possible to predict trends of patterns during an observation, supported by advanced analytics approaches. This could be useful to avoid failures, predict wear, avoid KPI value deviations, or reduce downtimes. Automatic adaptations in control systems in case of a (predicted) known system anomaly could automatically avoid a failure through specific control adaptations (e.g. reduce the load of a manufacturing system in case of an overloaded system). In general, monitoring and observation could increase the visibility of smart manufacturing systems. |
| Diagnosis | Big Data analysis could support the diagnosis (identification of causes) of anomalies, failures, etc. based on (historical) data analysis, which accelerates the diagnosis process, and results can be saved as a pattern, usable for future automatic prediction and detection, also of similar problems and also in similar machines. |
| Decision | Results of Big Data analysis could provide new additional information for operators or intelligent modules of a system, usable as decision support, or even decision making. |
| Optimisation | Manufacturing system effecting causalities could be identified through Big Data analysis, useful to optimise the related system. An example could be the optimization of KPIs, as the production time per product or the energy consumption per product. A further exploitation could be maintenance plan/scheduling optimization and individualization for manufacturing systems, considering the load, the ambient conditions, and the usage of a system, e.g. to reduce service costs. |
| Feedback | Product feedback as additional Big Data source for an analysis could provide additional hints for potential product and manufacturing system/process improvements, in order to get better results. |