There are two ways to digitalize and network an analogue industrial plant. The classic route is replacement procurement: a new, smart machine can simply be integrated into networked production. As the average age of machinery in German companies is between 15 and 20 years, this is often not an option, as the necessary investment for new machines would be too high due to the long remaining service life [1]. The 'smart retrofit' of machines, i.e. equipping existing production facilities with embedded systems and sensors with the aim of digital connectivity, can provide a remedy here. With the help of small investments, a still fully intact plant can be further developed into a cyber-physical system.

The 'digitalization toolbox' methodology toolbox

Figure 1: The '5 Times Why' method from the Toyota Production System.

© Bosch Rexroth

In order to fully exploit the potential of smart retrofitting, a systematic approach must be chosen. The 'digitalization toolbox' approach is based on the classic triad of problem identification, solution finding and technical implementation.

The '5 Times Why' method(Figure 1) from the Toyota Production System is used to define the problem [2]. It provides an approach that filters out the true cause of the problem. The method represents the leap to the technical implementation of the solution. For this purpose, the required function (result of '5 Times Why') is divided into several solution modules for realization. This makes it easier to reuse the components for other problems [3].

After splitting the required function into several solution modules, requirements for the technical realization are collected and transferred into a solution sketch. In order to be able to operate the solution economically, an important part of this method is to test its utility [4].

Digitization solutions must not only solve a problem, but also include the user and their work task in the economic solution.
Following the methodology for the conceptual development of the use case, the technical and organizational implementation must now be considered in the context of technical elements and the design of the necessary IT architecture.

The methodology in application

Figure 2: Logical structure of the CPS system with edge device.

© Bosch Rexroth

In order to upgrade a machine 'smart', the question of measurable variables must first be asked in relation to the technical elements. These measurable variables must not be determined according to the watering can principle, but must look specifically at the problems of the root cause analysis of the '5 Times Why' method. The resulting requirement data for analysis must be mapped using sensors.

An edge device with gateway function and intelligent control is suitable for data acquisition and data transmission for higher demands. It is also important to mention the scalability of the system: if further use cases are to be mapped subsequently, the system should be expandable at any time without major additional effort and costs.

The basic prerequisite for data transfer is a network topology with an adapted security concept. The core elements are the edge device, the MES and ERP or an IoT platform. The edge device handles all activities relating to the machine. It receives orders from the MES and controls the machine via an integrated PLC. At the same time, the device receives the data from the newly integrated sensors and controls the corresponding actuators based on the knowledge gained from the data. Figure 2 shows the structure of the Smart Retrofit System.

The methodology will be explained using the example of an SMD soldering oven (surface-mounted device). The challenge in this example is maintaining a constant temperature inside the oven. The '5 Times Why' analysis has shown that, in addition to the heating element, the temperature is largely regulated by a fan in the housing. The aim of the use case must therefore be to detect the fault, for example failure or vibration, of the fan and start an escalation of the fault.

Using the '5 Times Why' method, three solution modules were identified and formulated for this use case:

• Detection of a fan fault
• Escalation of the fault
• Resetting the fault after the problem has been rectified

The realization of the superstructure

Figure 3: Physical structure of the 'Smart Retrofit' solution of the SMD soldering oven.

© Bosch Rexroth

The analysis of the utility of the solution modules has shown that the MES cannot trigger any further meaningful actions with the isolated information of the sensor value. The notification is therefore extended with information about the machine name, time stamp and Si unit of the measured value. With the help of the basic topology and the solution sketch, the structure shown in Figure 3 can now be realized.

In the given application, an order is started by reading in a data matrix code. The edge device gives the control command to the machine and starts the fan. While the job is being processed, the vibration sensor collects data that is read out to detect a fault in the fan (see solution module 1). If a warning threshold value is exceeded, the system automatically issues a ticket in the MES via a node-red workflow to inform maintenance of the error message. The necessary spare part, in this case a fan, can also be procured automatically via a ticket in the ERP system. The software of the edge device is adapted so that the error message can also be automatically canceled.

Literature

[1] Etz, D.; Brantner, H.; Kastner, W.: Smart Manufacturing Retrofit for Brownfield Systems, Procedia Manufacturing, 42/2020, page 327 - 332.

[2] Ōno, T.: The Toyota Production System. Campus-Verlag, 1993, Frankfurt/New York.

[3] Griss, M. L.: Software reuse architecture, process, and organization for business success in Proceedings of the Eighth Israeli Conference on Computer Systems and Software Engineering. Herzliya, Israel, 1997. page 86 - 89.

[4] Ergonomics of human-system interaction - Part 11: Usability: Terms and concepts (ISO 9241-11:2018). German version EN ISO 9241-11:2018, DIN EN ISO 9241-11, DIN Deutsches Institut für Normung e. V., Berlin, Nov. 2018.