Machine data plays a significant role in a production environment, as operators strive to constantly optimize their processes and increase efficiency while maintaining the required quality parameters. Recorded machine data provides comprehensive insights into the condition of the monitored machines and their components. This data can be used, for example, in operations for condition monitoring, predictive maintenance, quality control and automation. However, the collection, transfer and storage of machine data is still a key challenge due to various causes. Current condition monitoring approaches are often used alongside other data-intensive methods such as machine learning and deep learning algorithms. However, there is a pressing need to develop new concepts for discrete data analysis that enable accurate fault diagnosis and effective maintenance planning without relying on extensive datasets. This paper presents a discrete concept for resource efficient production system condition monitoring. This concept consists of a data processing pipeline where many data reduction steps are done. Using the collected data of a drive, a principal component in production, we showed that the reduction of the parameters to be collected and the aggregation of data does not necessarily change the informative value and consequently condition/state determination is still possible. The data reduction steps along a data process pipeline showed the potential to save resources. This should relieve the network and reduce storage and computing capacity.