The Industrial Internet of Things With Retrofitting to IIoT
Very few companies are able to build a perfect production hall on the greenfield and follow the rules of IIoT to the letter. Automation and software companies have also become aware of this issue and respond to it by offering products for upgrading, retrofitting and retooling.
It is rarely the case that a factory is completely redesigned, built and equipped just to implement digital workflows and to set up a network of machines. The typical German medium-sized businesses in Germany produce in old locations and use a very heterogeneous machinery — both in terms of age and machine manufacturers. This presents companies and IIoT service providers with major challenges, since the aim is to successively network all machines and digitize processes.
Often, there is a general assumption that SMEs would oversleep the digital transformation or find it difficult to implement it. Looking at Deloitte's study "Industry 4.0 in small and medium-sized enterprises" from 2016, a somewhat more differentiated picture emerges: 28 % of companies are already completely networked, but with 29 %, almost as many companies have no networked infrastructure. Maybe it depends on the industry or product –however, this question was not investigated in the study. In the end, however, there remain two big questions to be answered:
- How can the existing machinery be upgraded or converted in such a way that the individual machines can be networked with each other?
- How do you implement the task of "installing" digital processes in a time- and cost-efficient manner with the existing employees?
A major challenge — now and in the future — lies in the different "lifetimes" of mechanics and software: The mechanics in machines are designed to last many years and decades, while software and IT systems quickly become obsolete. In order to network an outdated factory or production line and digitize the processes, the machines must become intelligent and their information and data must flow into intelligent software that creates added value. Existing control systems must also be retrofitted or upgraded. It is therefore not surprising that a large market for retrofitting solutions is emerging.
Automation companies in particular have already discovered this gap and are now offering products that are supposed to enable "Industry 4.0 for retrofitting". One of the first ways to connect a machine to digital capabilities was probably the Pepperl+Fuchs Smart Bridge. With this adapter, older sensors can not only send process data to the machine control system, but also communicate via Bluetooth with the higher-level systems which then evaluate all relevant data. For this purpose, the adapter is looped into the wiring between the sensor and the machine control, from where it picks up the sensor data without any interaction and establishes a wireless connection to a mobile device. The adapter does not require its own power connection, it uses the power from the sensor line. The readout sensor data is displayed on a tablet or smartphone via the smart bridge app. Using the app, users can also change the operating parameters of the sensors connected via smart bridge. Since data exchange takes place via a Bluetooth connection, there is no need to install additional cabling or interfere with the controller.
SKF's Enlight Quick Collect works completely without installation, but it should also make the work of maintenance engineers easier. The product consists of three components: a portable sensor, a mobile device and the Quick Collect app. The sensor collects machine data such as vibration and temperature values and transmits them to a smartphone or tablet using conventional mobile technology. The Quick Collect app installed there performs a basic analysis of the measurements in real time and informs the user immediately about possible problems. In addition, the app saves the findings for in-depth analysis and enables the forwarding of relevant information.
Making Machines Talk
The Bosch Rexroth IoT gateway also works without any intervention in the control system. Nevertheless, it should also be possible to network decades old existing machines with little effort. The inconspicuous box consists of the embedded control Indracontrol XM, which executes the software of the gateway. The architecture is based on open software with Linux as the operating system and a Java virtual machine that enables the deployment of Java applications. Process data such as temperature, pressure, vibration or power consumption can be recorded in real time via a sensor package from Bosch. Thanks to integrated software apps such as the device and processing app, users can easily customize the IoT gateway via a web-based platform.
What does the IoT Gateway do? It collects various sensor and control data and forwards them to a central database for analysis. This also works with older controllers without having to re-program them, as well as with third-party controllers from Beckhoff, Siemens and Rockwell. In order to obtain information or recommendations for action from the data, they are now connected to existing IT systems such as MES or special evaluation software. Thanks to its high connectivity, the IoT gateway can communicate with many higher-level IT systems just as easily as with cloud platforms.
Bosch Rexroth demonstrated the implementation of such a digitization application at Bosch Rexroth's Lohr plant using its own engine production facility and divided it into five exemplary steps:
1. Defining influencing factors: In order to optimize a machine, for example, certain parameters are important which influence the product quality.
2. Selection and retrofitting of the sensors: The information required for monitoring the machine must be collected by sensors in order to avoid machine failures.
3. Connecting the sensors to the IoT gateway: The collected sensor data is transferred to the higher-level IT systems via the gateway.
4. Connection of the higher-level IT systems: The processing app is used to determine the data transfer.
5. Exploit potential for improvement: New measures can be derived from the data to bring about improvements in production.
How Machines Learn OPC UA
ZNT-Richter has developed another option for retrofitting communication interfaces. Its Process Automation Controller (PAC) is an open platform that connects the control level with the MES or ERP. In addition to data collection and aggregation, PAC can also perform automation tasks at the workspace level.
These include cross-machine production control as well as the integration of process control and optimization, machine program management or material flow control. The platform also relieves the MES level of automation processes in real time, because connected machines can be decoupled from the MES if necessary. This in turn enables flexible start and stop cycles for individual plants without shutting down the entire production process. The elaborate evaluations can be transferred by PAC, raw data can be written directly into a process database.
Machines Learn a Foreign Language
OPC UA is becoming mature as a communication standard for networked factories. But not every machine speaks this language. Researchers at the Fraunhofer-IOSB have investigated this problem in cooperation with Codewerk. The result is the Plug-and-Work Cube retrofit box consisting of the Siemens IPC "Nanobox" and an assistance system that allows the production equipment to be described without modelling know-how. The Nanobox contains an aggregating OPC UA server with Automation-ML as information model and can be automatically coupled to Siemens PLCs and underlying OPC UA servers. The assistance tool is a Windows application integrated in the official Automation ML editor as a plug-in. It can be used to create the configuration for the aggregating OPC UA server.
Small Change, Great Relief
Sometimes, however, the problems already start at simpler areas: How can I obtain diagnostic data from a control valve or how do I find out when my products are ready to be dried — without having to install a complete automation structure? The sensor and safety technology manufacturer Turck has been developing individual solutions. The control valve is part of a rotary actuator. It's all about whether the valve is open or closed, i.e. the time from the moment the valve is activated to the signal "Valve is open". Turck has equipped a TPM module with the function "Field logic controller". This function measures how long it takes the valve to open. The data provides several relevant pieces of information. For example, users can determine whether maintenance is necessary. The analysis data are displayed in curves that can be evaluated directly on the module. An additional control unit is not required.
Another simple solution is the monitoring of the drying time in connector production. The manufacturer had not automated the drying room; when a load of connectors was ready, the employees took over. Turck has only installed a signal lamp, RFID read/write heads and a small module. The task of the installation is to register when a trolley arrives and to identify it, to monitor how long it remains in the station and to compare these values with the programmed waiting time. This period depends on the drying time required to ensure the quality of the connectors. From the signal light at the terminal, the user immediately sees when a trolley can be taken out of the room. This does not require a control system or a complete automation structure.
A Retrofitted Machine Tool
Contact Software and the Fraunhofer-IPK have undertaken a more extensive conversion, with more extensive results. The team carried out a retrofit on a machine tool. However, it was also a rebuild, the machine is now used as a test bench. It is based on a MEMS-based sensor network that collects and evaluates data at runtime. The machine was also connected to the Contact Elements platform. As a result, operating data from live monitoring and the manufacturing status of the test bench and the test specimen can now be combined in a digital twin. This is supplemented by a digital CV file which documents the manufacturing status and digitally supports test sequences. For the measurement data analysis, further software was imported.
In the final analysis, a retrofit requires networked sensor technology and integrated communication. Automation manufacturers and research institutes have finally taken these topics to heart, and they continue to work on feasible retrofitting solutions for existing systems.
This article was first published by MaschinenMarkt.