The Road to
SMART FACTORY

The transformation of a factory into a smart factory can be illustrated using the 6-step model by Prof. Dr. Schuh et al. be described in an understandable way. The respective stages can be viewed and evaluated individually, but it is necessary that the previous stages have been carried out, these build on one another.

In the first stage of digitization through the lean transformation to the autonomous factory, the data must be recorded digitally. Existing manual measuring systems must at least be implemented as part of a BDE (operational data acquisition), better an MDE (automatic machine data acquisition). The more automation-capable sensor technology is used here, the more digitization can be advanced.

The second stage relates to the networking of the sensors and data acquisition that have been set up. Necessary protocols are to be defined, network infrastructures created, and databases set up and connected to store and effectively transform the data that arises. With the completion of this stage, the transformation step of "digitization" is completed, but both stages do not yet provide any advantage, but would only cause costs on their own.

By making the collected information available in the third stage, the intermediate goal of the "digital factory" has been achieved, all currently occurring data and their historical processes are to be visualized. Ideally, there is a "digital shadow" or a "digital twin" behind this data. These are already suitable for running simulations on them, for example.

The visualized and meaningfully structured data, especially in the context of a "digital twin", are suitable for carrying out data analyzes on them. Correlations between different data series, anomalies within the series, as well as further trends and patterns can be determined. The data analysis as the fourth stage is a tried and tested means of precise analysis of the past and provides important information for improvement and gives indications of sensible starting points for changes and value creation potential.

The fifth stage in the context of "predictive analytics", i.e. the forward-looking analysis of the existing data, provides an outlook into the future, especially for individual machines or their sub-components. Outgoing from the data analysis of the previous level, the current data can be compared with existing situations from the past and thus provide recommendations for action, e.g. for predictive maintenance, repair or replacement. The aim, particularly in the context of maintenance, is to make maximum use of the remaining service life of components and machines while at the same time minimizing the risk of an unplanned downtime.

After all data has been digitally recorded, networked, visualized and analyzed, and an outlook into the possible future has been made possible, the prerequisites for feedback loops within complex processes across multiple machines are in place. In the "Smart Factory", neither the processes nor the machines involved are specified, but the target goal is defined. The autonomous factory, which represents the sixth and final stage, optimizes the process and the machines required for it itself. Neither higher-level planning nor human intervention in the process flow is required.