5G adoption is expected to take off in a big way in 2023. In late January, Vodafone made its long-awaited 5G Standalone (SA) technology available to trial customers across the UK. But enterprise users are also keen to explore the prospect of ultrafast connectivity for localised environments, in office buildings, factories, warehouses, hospitals and transport. Forty-five per cent of UK businesses are planning to make investments in 5G by the end of the year.
A great promise of 5G is the ultra-low latency enabled by mmWave (a theoretical 1ms and under, compared to 20-30ms with 4G), which can provide businesses with near-time data from sensor-equipped devices to boost productivity and reduce costs. With that in mind, a welcome breakthrough came with Qualcomm’s completion of 3GPP Release (Rel) 17 last March, which addressed earlier range and stability issues, thereby supporting new deployment scenarios.
A survey carried out by Nokia shows a high level of confidence in the potential of 5G to deliver enterprise cybersecurity, cost, efficiency and sustainability goals. Nokia also found that over 50 per cent of 5G adopters had already seen total cost of ownership (TCO) reductions of 6 per cent or more from their investments, with 29 per cent experiencing a more than 10 per cent reduction. Eighty per cent expect to achieve a return on investment (ROI) within six months.
All the indicators suggest, therefore, that the 5G age is nearly upon us.
New radio
Release 17 will enable developers to focus their efforts on standard low-power devices, operating on the 5G New Radio (NR) radio access network (RAN) at bands up to 52GHz. This paves the way for a new era of low-cost sensors that can operate on lower power 5G NR modems. NR positioning targets 20-30cm location accuracy, which is perfect for factory automation. Impressively, 5G networks allow for up to one million sensors per square kilometre.
With 3GPP’s new RedCap tier, meanwhile, the support of a single antenna branch also accommodates more compact device form factors, which should help business users to incorporate popular wearable tech like smart watches.
Latency of seven milliseconds and under would enable businesses to use virtual reality (VR) and augmented reality (AR) for training, design and real-time communication between geographically dispersed teams, reducing the need for travel to physical meetings. Formula-E battery manufacturer Hyperbat – which is a joint venture between Williams Advanced Engineering and Unipart – has deployed a 5G Mobile Private Network (MPN) for just this purpose.
Hyperbat makes use of the world’s first 5G virtual 3D engineering model to optimise the connection between the company’s Oxfordshire-based design team and its factory in Coventry. Masters of Pie’s real-time collaboration engine, Radical, meanwhile, is integrated directly into CAD software, giving nearly instant access to all relevant data.
According to information issued, the consortium’s proof of concept addressed all the challenges of extended reality (XR). Teams are equipped with Qualcomm’s Snapdragon XR2 5G reference head-mounted displays (HMDs), which are untethered and come with seven cameras, providing a 2k by 2k display for each eye, six degrees of freedom, and a state-of-the art eye tracking feature. Thanks to the SDX 55 modem, the devices support both mmWave and sub-6GHz.
A technique called split rendering enables perception-based activities (including eye tracking) to be carried out locally by the headset, while the more computing-intensive actions are processed in the cloud and then streamed to the device by Nvidia CloudXR.
Production line of the future
Auto industry analysts reckon that up to 70 per cent of production costs can be eliminated with improved decision-making in the design phase. Katherine Ainley, CEO of Ericsson UK and Ireland, predicts a transformation of the shop floor: “We expect the production line to become even more efficient. We can reduce the number of wires and maybe even move to a wireless production line.”
Ford has adopted similar measures at its E:PrimE (Electrified Powertrain in Manufacturing Engineering) facility in Dunton, Essex, calling on Vodafone to custom-build an ultra-low-latency 5G MPN. This enables secure, real-time handling of data generated by the welding of batteries, a process involving around 1,000 welds and up to 500,000 pieces of data per minute. The new MPN helps Ford to speed up assembly and ensure greater precision.
Traditionally, the company’s equipment was mostly fixed and hard-wired. But that is set to change. “There is likely to be a lot more customisation in cars,” according to Chris White, manager of the 5GEM Project. White says 5G should help Ford to readjust as the industry evolves. He has been looking into the use of untethered robots and AGVs to complete repetitive tasks, and the introduction of remote assistance via AR.
Ericsson claims that its customers have achieved cost savings of up to 20 per cent by using 5G technology in industrial settings. These savings are typically realised through increased efficiency, productivity and asset utilisation. But vendors are also targeting maintenance, a massive cost centre for manufacturers, as broken machinery causes production delays that can set companies back many millions.
At the same time, 5G-powered predictive maintenance helps prevent failures before they occur. The UK’s very first live private 5G factory installation saw BT help engineering firm Worcester Bosch to successfully raise factory productivity by around one per cent, via the use of autonomous robots to transport products and materials. Thousands of sensors were used for health monitoring and collision detection, enabling the company to minimise breakdowns.
Meanwhile, in Germany, Mercedes-Benz has built a 5G-enabled auto plant, Factory 56 in Sindelfingen, which optimises data linking and product tracking on the assembly line. In Munich, MTU Aero Engines teamed up with Fraunhofer IPT in a bid to harness 5G to more quickly detect manufacturing failures in its production of bladed disks (BLISKs) for jet engine turbines. It
is a process that is notoriously difficult to monitor.
Regarding the latter process, the current high rework rate of 25 per cent is largely due to vibration and ‘chatter’ during milling. The emerging industry trend towards thinner blades with increased aspect ratios, meanwhile, means the newer blades are even more flexible and prone to vibration.
However, MTU’s engineers can now tell within milliseconds when it is necessary to alter their machine parameters. They have reduced rework considerably, saving €3,600 per BLISK – or €27m per factory per year.
Standalone versus non-standalone
The widespread adoption of industrial sensors in manufacturing represents a £2bn yearly contribution to UK GDP, according to Mark Stansfeld, chair of West Midlands 5G.
While the telecoms industry is enthusiastic about the uptake of standalone 5G networks (ie, those based on a pure 5G core), non-standalone networks have also delivered tangible results for industrial companies looking to optimise their processes by building on top of existing 4G infrastructure.
For instance, Edwin Lowe Group’s AE Aerospace has built a ‘glass factory’ in Birmingham that uses an NSA 5G network to move and analyse data from its inspection devices and tracking sensors in real time. It can monitor components as they move around the shop floor, get maintenance updates and detect faults.
The system’s ‘Vision Intelligence’ camera system takes hundreds of images of a single surface on a component, quickly identifying scratches and other imperfections. Meanwhile, the network is based on a 4G LTE core, but 5G is vital for moving the large file sizes.
As a manufacturer in subcontract machining with clients around the world, this fault-detection feature has proven a valuable addition: “It has improved the quality of our products and we’ve been able to get them out at a faster rate,” according to Steph Potter, AE’s continuous improvement officer.
Other companies are looking to make use of the network slicing techniques made possible by SA 5G. Slicing enables companies to create multiple segmented virtual networks, each of which can be assigned different characteristics. Factories are thus able to provision extra network ‘slices’ on the fly, thereby supporting shifts in the volume of production. Manufacturers only pay for what they need, creating longer-term savings.
Challenge landscape
Currently, digitalisation is led by large manufacturers in the high-value manufacturing space, particularly aerospace and defence, rather than SMEs or manufacturers from lower-margin sub-sectors. And there is, frankly, fatigue around popular terms like ‘industry 4.0’ and ‘industrial digitalisation’ as companies do not know where to start. They are wary of marketing hype, and many have delayed equipment upgrades as they rein in discretionary spending.
Vodafone has cautioned that SMEs in particular may fall behind on 5G adoption because they lack the deep pockets of larger corporates. But the operator’s recent SMEs Like Me study also warned that large technology vendors shouldn’t treat SMEs as a homogeneous group, since they are as wide-ranging and complex as many of the obstacles they face.
It is convenient, then, that a number of smart factory demonstrators have been installed across the UK, including the Siemens initiative at the MTC in Coventry. Added to this is the centre for Aerospace Manufacturing at the University of Nottingham, and the Bosch Rexroth smart factory facility. Vodafone itself has established ‘V-Hub’ to provide one-to-one guidance to SMEs, and late last year installed a demo 5G MPN at Coventry University’s Uxplore hub.
The hub invited businesses from across the region to examine 5G’s potential at close quarters.
