No Industrie 4.0 without any semantics!

It is clearly a challenge to connect interlinked machines and equipment with a higher-level IT system because the lions’s share of the configuration of the IT system is carried out manually. This is owing to the fact that many of today’s manufacturers outsource production processes and lines to specialized equipment manufacturers. The mechanical and plant engineering sector is marked by medium-sized enterprises, which is why manufacturers obtain their equipment from multiple providers. Consequently, the plants contain heterogeneous equipment, consisting of a variety of controls, multiple communication protocols and/or field bus systems, etc. Unlike the PC world or consumer electronics, there is no plug-and-play with automated configuration processes based on USB connections. The connection between plant control and the higher-level IT system is largely rigid and configured in a plant-specific way. This results in immense configuration efforts for the initial start-up and for each adaption of the production equipment to new framework conditions.

The difficulty for medium-sized manufacturers of machinery is that they increasingly have to comply with their customers’ specifications with regard to ‘MES connectivity’. For this reason, it only makes sense to standardize the designations of their own MES-related signals if the MES is fed from a ‘meaningful’ data repository. In other words, the meaning of the individual data points is described in a generally comprehensible and machine-readable way.

We have to distinguish between two practical cases of self-description:

1. During the development process of the machine, the engineering process includes a self-description in the form of a model.
2. For an existing machine, a self-description is to be worked out only after it has been set up, on the basis of the data points already available in the machine. This case is more difficult, which is why we have developed our PLUGandWORK Cube.

In line with VDI 5600-3 (Guideline of the Association of German Engineers) a data point generally stands for the data that has to be transferred between the machine and the ‘outside world’ including “unique designator, meaning, potential synonyms, the data format and the information whether it is mandatory or optional contents”. The self-description also includes interfaces and access authorizations.

So if the manufacturer of machinery and equipment uses planning systems that generate data during the engineering process (mechanical and electric plant engineering and controller programming) and which are required for the customization of shopfloor-related IT systems, this data is included in the self-description. This manufacturer-specific data can be enhanced by context information, for example from the tools of the higher-level electrical planning or the material flow or layout planning. In case the layouts are already available in a ‘structured form’ – i. e. their elements have been stored as objects that can be addressed individually, the layouts can later provide targeted elements for plant visualization or process control images. Our “PLUGandWORK for material flow Project” with Gebhardt Fördertechnik provides an example for the generation of the HMI.

Using a similar approach as a USB interface, plant components can exchange data independently. The advantage is that the controlling software can identify new or modified system components quickly and easily. All the information is transferred to the production process for automated integration. The motto and method are called “plug and work”, which is based on the use of open standards already in use in today’s industry. For this reason, no additional interfaces or drivers have to be programmed or customized to the production equipment.

Taking machines or machine tools into operation is one of the most time-intensive steps, the time and cost of which are often hard to anticipate. For a modern production line, for example, the manufacturing and assembly of the individual machines can take up to six weeks. Taking the overall installation into operation may be almost as time-intensive. Depending on the project layout, this takes place on the manufacturer’s premises or on the customer’s construction site. The lion’s share of this time results from the fact that the machine manufacturer has to identify and to store the component characteristics manually in the machine control, for example ball screw pitch errors (see, for example, left side of the figure, taking the form of a pdf file or on paper owing to the data transfer). When components are integrated into a machine tool, specifications such as the slope of the ball screw drive, the bearing distances and the electrical indicators of the main spindle have to be stored in the machine control by manual efforts. Owing to this manual share, this is a time-intensive and error-prone process. To ensure the required process capability, component-specific errors have to be compensated for. For example, the geometric errors in ball screw drives are identified by means of a ballbar test and the resulting compensation values are stored in the control system. In the case of a main spindle, the shifts of the processing tool resulting from temperature-related impacts have to be identified by appropriate external measuring devices so they can be compensated for.

For example, there is a 7-digit numeral code on the screw nut to identify ball screw drives. This laser-engraved code is required to connect the ball screw drives with the start-up data of the individual components. To this end, pitch error protocols, frictional torque curves, rigidities, geometric data and even test certificates for the respective ball screw drives are necessary. This data is forwarded to the customer in various data formats and separately from the ball screw drive. The data from the test measurements at the test stands is generated in multiple data formats. This results in high manual efforts, which is driving the costs for the start-up of machine tools.

In addition, the start-up of expansion systems such as gripper systems used for work piece manipulation is still time and cost-intensive today. In this case, for example, the potential motion ranges have to be stored in the higher-level control system. In addition, the positioning accuracy is stored in the repository for high-precision positioning processes. A machine tool can be equipped with various tool magazines. When taken into operation, the number, position and condition of the individual tools has to be stored manually in the control system, which can amount to up to 100 units in the case of large magazines.

In addition to the standard-based communication of machines and components, PLUGandWork includes the generation of process control images for visualization systems (HMI) on the basis of the self-descriptions. By means of AutomationML language constructs, geometric and topological data is converted into the visualization during runtime.

Even short-term modifications in the hardware are considered in the model and immediately integrated in the new HMI. This will briefly be described using the example of standardized, flexible and adaptive conveyor technology modules by Gebhardt GmbH. The precondition are modular conveyor system units for containers, boxes and palettes, which are combined to form the conveyor system using modular design. Currently, there are 14 conveyor modules in total, each of which belong to a pre-defined product series. The production series reflects various dimensions (length and width) and possibly angles according to a pre-defined size-graduation. To ensure that the 14 conveyor modules and its production series can be combined to form various conveyor systems, they take a modular outward appearance. They can be combined with each other as desired. To this end, they have been equipped with mechanic and electric interfaces.

Hardware and software have a modular design, which means that each module of the conveyer technology contains its own controller including the related model of the corresponding module. The modules communicate with each other using Ethernet. To generate the process control images in multiple visualization system, topology data, topography, structure, geometry and IO connectivity are included in the self-description.