The UK’s leading telco BT has made the bold claim that only private 5G can deliver smart and flexible manufacturing, as part of the industrial IoT (IIoT).
BT, via its EE mobile arm, has been lead technology partner for the Worcestershire 5G Testbed (W5G), which now claims to be the UK’s first live private 5G network. Using equipment from Ericsson at Bosch’s factory in the western city of Worcester, and the nearby Malvern Hills Science Park, it employs network slicing technology from Zeetta Networks for use with private and public networks currently under construction at the National Composites Centre (NCC) in Bristol.
Building on the first set of UK 5G testbeds and trials, the expansion of the Worcester project will provide a valuable testbed for the IIoT because it will support many of the relevant technologies and concepts, including sensors, wearables, data analytics, robotics and augmented reality. It will also incorporate localized edge and multi-access edge (MEC) technologies, as well as some cloud-based computing. The stated purpose of this demonstrator project is to establish a blueprint for Industry 4.0 in the UK with wide-band cellular communications as the springboard for “intelligent, dynamic and fully automated” manufacturing processes.
BT talks grandly of how the project will develop the understanding and expertise needed to “optimize private networking, edge computing, and data analytics for the fourth industrial revolution”. Steam power was the first revolution, electric power the second and computer driven automation the third, so the fourth is really an extension of the third rather than a fundamental shift. If there is any revolutionary ingredient it is wireless communications on the back of 5G, which has been held back so far by lack of resilience and insufficiently low latency.
As for the claim that private networks are essential for smart manufacturing, that is only half right. It is true that real time control in manufacturing requires the ultra-low latency that can only be delivered from the network edge. It is also true that security is critical, given sensitivity of some of the data and above all requirement for insulation from external tampering. Furthermore, enterprises in this sector are attracted by the idea of being in total control of their networks and not having to share them with anyone else. That they believe will improve reliability and ensure QoS without risk of bandwidth being consumed unpredictably by other users or organizations.
However, 5G has been designed precisely to support disparate use cases and multiple entities via network slicing. It is true that security is a concern, but no more so than for the automotive sector where access to public cellular networks is essential for both entertainment and real time traffic updates in navigation for example. That sector will have to reconcile need to protect safety critical systems with access to public networks, and similar defences could be employed in the IIoT.
The point is that while some manufacturing operations will indeed by well-served by private 5G networks using spectrum not available in that locality to anyone else, there are some instances where wider connectivity will be essential. Indeed, the GSMA published a white paper identifying five sub-categories within the overall IIoT use case family, of which three would require access to public 5G networks.
Of the two cases that could be served exclusively by private networks, the first is real time process control, involving time-critical closed loops requiring ultra-low latency, similar to the BT UK demonstrator.
The second IIoT use case involves less time-critical communications but which can still be served by local communications and perhaps generate sensitive data. This includes various on-site production and logistic processes, quality control and data capturing for subsequent analysis. Such processes either entail collection of data for processing later, or involve human interaction that does not call for sub-second latencies.
However, the GSMA’s third IIoT use case family addresses remote control of factories, which by definition requires wider area access over a public network. It is true such access could occur over a fixed line network, but that would still bring the same issues over security and latency. Furthermore, wireless would still be important for enabling full mobility among mobile elements within an automated, unattended or “lights-out” factory. As the GSMA pointed out, for this use case the public wireless access network will take a role in the end-to-end communication between remote workers and the factory.
This is considered one of the most exciting IIoT use cases that 5G will enable or enhance. There have been isolated examples of unattended factories for some years, but these have been confined to products that could be made readily with technologies available at the time and without full mobility on the shop floor.
One of the earliest examples dates back to 2001 when Japanese robotics firm FANUC started operating a lights-out factory where its own robots were building themselves in unsupervised runs lasting for up to 30 days at a time. This demonstrated the cost savings not just in human resources but also associated environmental control, such as air conditioning and heating no longer needed for staff comfort. More recently, Philips started making electric shavers in the Netherlands in a largely unattended factory, just with quality assurance workers to oversee the final stages of the process. There is also use of partially unattended operations to manufacture some integrated circuits, again with some human oversight.
With 5G there is hope that even quality assurance can be automated, or at any rate performed remotely, while extending unattended operations to other manufacturing processes, including automotive and larger electronic appliances. There is also potential for enlarged scope of remote operation, for example in preventative maintenance and increased accuracy for better yield and quality.
There is also interest in application of augmented reality (AR) for various remote scenarios, including overlay of accurate operational detail for directing precision machinery, and for repair without having to visit the factory. This is very much in the spirit of the times, reducing the number of skilled engineers needed and increasing their efficiency by cutting travel time.
The GSMA also identified two other rather more futuristic use cases where public 5G would play a role. The first is what it calls inter-enterprise eco-system communication, or ‘design anywhere, produce anywhere’, enabling greater specialization within manufacturing where multiple production sites and other players in the chain, such as suppliers and logistics companies, interact in real time. This would extend and diversify the JIT (Just In Time) concept to coordination of assets and activities across and even between whole supply chains.
The remaining IIoT use case where public 5G would figure is sometimes called connected goods, where profiles of products are maintained and updated during their lifetime from data collected in the field, for various applications including predictive maintenance and computer aided design. Currently such information is usually not available until it is too late, or has become redundant, especially from the design perspective.