Two prototypes have been built based on the 4th generation of the DMC 80 FD duoBLOCK machining centre. One prototype will be put into volume production in a precision bearing cell at the Schaeffler plant in Höchstadt. The other was unveiled on the DMG MORI stand at the EMO 2015 exhibition.
Schaeffler is pursuing a specific digitalisation strategy, with the aim of providing data from a diverse range of processes using sensors, networking and analyses in order to offer its customers real added value.
As a user of machine tools, this Industry 4.0 initiative has a direct relevance to Schaeffler’s own production. Schaeffler’s activities in the field of machine tool digitalisation provide the whole sector with the potential to convert opportunities and overcome challenges by close collaboration between manufacturer, supplier and end user.
Bearings as data sources
Bearings are key components for machine tool performance, as they are critical not only for the functional capacity of the machine, but also for the quality of the work piece. Information about the current condition and the future behaviour of components will become an important source of information for machine operators. Existing sensors can be used or suitable equipment retrofitted. At times, it is advantageous to integrate the sensor directly into components, as this is often the only point at which some parameters can be determined.
In this project, the two prototypes have additional sensors integrated into almost every bearing position relevant to the machining process. This enables vibrations, forces, temperatures and pressures to be measured, which provides valuable information about the condition of the machine. Making a machine ready for ‘4.0 production’, means assessing the measurement data, saving this data and making conclusions from this.
In order to make all the measurement data accessible, the machine is provided with an internal network to which all additional sensors, actuators and evaluation units are connected. A gateway provides a link to a Cloud-based platform. In order to ensure that data can be exchanged with the machine control system, Profibus is integrated into the PLC for time-critical and process data. The OPC UA protocol is used to access further information from the human machine interface (HMI). The data from the machine is saved locally in the gateway and copied into the Schaeffler Cloud. This ensures the machine’s data history is available without having to connect to the network. Calculations can be completed in the Cloud via web services or apps.
Big Data and data analysis
Analysing large volumes of data has assumed a new significance compared to existing data analysis, which principally has a 1:1 ratio as far as output is concerned. This assumes that, in addition to the actual measured values, trends occur in a sufficiently large number of measured values/data (Big Data) which can be correlated with other data. This provides a new level of quality in terms of what it can reveal with regards to, for example, the condition of the bearings and therefore the condition of the machine (data-based added value).
These data trends can be automatically recognised using suitable software algorithms and any necessary recommended actions taken. Decentralised functional units are required, which can operate both autonomously and as an integral part of the network. This allows local intelligence to assess the data locally. Additional evaluation that requires more computing power can be retrieved from the Cloud. An analytical evaluation based on data from all connected machines can be carried out in the Cloud rather than locally on the machine.
In this way, Schaeffler is designing a horizontal network along the value-added chain, similar to the vertical integration of sensors into the Cloud, in order to learn how complexity and requirements for products and services can be accommodated in production.
Integration into production
The possibilities offered by digitalisation are not limited to production machines. The manufacturing environment can also benefit from a continuous flow of data. This avoids the implementation of isolated solutions that are likely to require manual intervention. Vertical integration is also required to connect to the Enterprise Resource Planning (ERP) system for automatic order processing.
An important element of ‘Big Data’ is the unique identification of individual components. For this, a marking unit is integrated to the machine, which provides each component with a unique code using a Data Matrix Code. This remains with the components throughout the manufacturing process and forms part of the unit ID. This provides traceability and enables a component’s history to be analysed.
Determining the forces on the tool centre point (TCP) allows further optimisation of machine loading, as well as the process itself. Displacement at the TCP due to loads that occur during machining can be determined using a mathematical model and potential correction measures can be fed back to the control system in real time. The actual machining forces that occur can be determined in advance using machining simulation software. These form a nominal value that may not deviate outside a predetermined range, as this would indicate an unacceptable condition.
In addition to measuring actual energy usage and assigning this to each stage of the machining process, it is also possible to determine future energy demands through process simulation. When combined with existing data, a more accurate prediction of energy requirements can be produced, which allows demand-driven purchasing of energy, as well as production planning by minimising energy peaks across the business.
Condition of the machine
The condition of the machine is recorded using traditional vibration monitoring techniques. The condition of the lubricant is also measured and evaluated at different points. Demand-driven lubrication guarantees functional capacity, as well as the careful use of resources without affecting machine performance. It is also possible to predict the future condition of the bearing.
Load collectives in the machine can be clearly understood by classifying the machining processes. For example, a nominal residual service life of the bearing position can be calculated online using the Schaeffler bearing calculation software BEARINX via a web service. The aim is to simulate planned machining tasks and the resulting (expected) operating life of individual components in order to control production in such a way that essential maintenance can be planned in advance to maximise machine availability.