This contribution gives a brief introduction to general aspects of “Industrie 4.0”. Besides basic strategies to
improve the added value and flexibility of a production, challenges of the transformation, which have to be
overcome by the companies, are shown. The commissioning of production machines gains more significance in a
dynamic production of a smart factory, so that in consequence the automation of the commissioning would bring
significant advantages. Current fluid power systems are not excluded, since most steps of the commissioning are
still done manually by the technician. By analysing the integration of a linear electro-hydraulic actuator into a
production machine, limitations and problems of current systems are identified and related to the field of fluid
power. The analysis of possible solutions is leading to methods and modern information and communication
technology, introduced by the “Industrie 4.0”.

We present a service for Predictive Maintenance in which existing machine data from control units or data from
retrofitted sensors can be acquired from industrial machines by various gateway solutions. These gateways
preprocess the data onsite and transmit it securely to a cloud-based Big Data system without impacting the
production process of the industrial machine. Additional servers run Machine Learning algorithms to analyze the
incoming data and generate data-based models representing the machine behavior. Results from existing
applications show that significant benefits can be created for our customers and that Machine Learning
algorithms demonstrate superhuman performance in detecting anomalous machine behavior.

Industrial Internet of Things (IIoT) and Industry 4.0 are very popular buzz words today. The „me too” factor is
pretty high and attracted companies are faced with an overwhelming market of data management solutions. But
despite the large amount of data that can be collected from industrial facilities, the real benefit is behind
colourful graphics and charts. To get there, the data provided by the connected components of an IIoT capable
system has to be analysed and put into context. So, the question is not what can be done with all the collected
data but how to generate useful information.

Today buzzwords like “smart machine” and “intelligent component” dominate the discussion about digitalization in the fluid power domain. However, the engineering fundamentals behind the words “smart” and “intelligent” often remain unclear. A common and target-oriented discussion needs transparent approaches including the applied technical system understanding. Therefore, this paper presents new concepts of soft sensor networks which allow the aggregation of information about fluid systems from heterogeneous sources. Soft sensors presented in this paper are physical models of system components that ensure transparency. Soft sensors and soft sensor networks are applied on exemplary hydraulic systems on three different levels: (i) the sensor level, (ii) the component level and (iii) the system level.