A reference architecture that takes into account the interaction of 5G and IIoT comprises five levels: Sensor level, network level, middleware level, application level and business level. In an IIoT system, sensors and actuators collect data that is transmitted via various communication technologies and processed in the middleware layer before being used for various applications in the application layer. The business layer manages the entire system and requires the development of better business models.

The URLLC and eMBB application profiles provide the basis for the following 5G IIoT technologies:

- Network Function Virtualization (NFV) enables the creation of virtual versions of network hardware that are easier to manage and update. It replaces dedicated hardware with off-the-shelf servers, making networks more agile, robust and cost-effective.

- With Massive Multiple-Input Multiple-Output (MIMO), several antennas and transmitters are used on the receiver side. This improves the data throughput, increases the maximum number of end devices in an area and increases the range of the system by superimposing the signals.

- New Radio Frequency (5G NR) is an access technology that operates in a range of low, medium and high band frequencies from below 1 GHz to 24+ GHz. 5G NR offers a function to provide the necessary bandwidth in the system to enable very fast data transmissions.

- Mobile Edge Computing (MoEC): This technology is seen as an alternative to the current centralized cloud as it provides resources close to edge IIoT devices. MoEC aims to reduce network congestion and speed up response times.

- Non-Orthogonal Multiple Access (NOMA) enables high spectral efficiency with limited spectrum. Simultaneous transmission of data from different sensors on the same resource block is possible. Furthermore, interference is reduced and spectrum utilization is increased.

Best practice in the context of production

The interaction of IIoT assets using 5G is being tested in the Learning and Research Factory of the Chair of Production Systems. For time-critical applications and information processing on the store floor in particular, 5G offers great potential in terms of connectivity, latency, transmission speed and quality, which emerge from the 5G application profiles URLLC, eMBB and mMTC.

To evaluate a 5G network in such a scenario, store floor assets are connected via a 5G wireless network according to the reference architecture model. These include various programmable logic controllers that monitor sensors and actuators. These are equipped with a 5G module for integration into the network and configured according to its communication protocols. Communication between the systems takes place via the 5G base station, which is connected to the OPC UA server and the SQL database via LAN cable. At the middleware level, the data is consolidated in this database and further processed at the application level by a Manufacturing Execution System (MES) and other applications for visualization purposes.

Furthermore, mobile and airborne robotics applications with 5G are already being implemented in the learning and research factory. In tests with ICMP packets, latencies of 10 to 15 milliseconds and data transfer rates of 300 Mbit per second in the downlink and 130 Mbit per second in the uplink are achieved, which is satisfactory for this type of application. However, the latency time for applications with programmable logic controllers should be as low as possible in order to be able to compete with other means of communication.

Gustavo Barros, Maximilian Bega, Daniel Gorsek and Bernd Kuhlenkötter from LPS, Ruhr-Universität Bochum/ red