Enterprises roll out private 5G while standards, devices, coverage evolve

5G promises dramatically faster speeds, lower latency and increased capacity than today’s 4G LTE technology. It’s also a potential alternative to Wi-Fi with greater reliability and performance in industrial settings.

Coverage, standards, and device support still aren’t where they need to be for full deployment, but that isn’t stopping some companies. Early adopters are rolling out 5G initiatives before all the pieces are fully in place, in order to reap immediate benefits and get a jump on the competition.

The U.S. Veterans Administration is one of those early adopters. It’s using Microsoft’s 5G-enabled HoloLens augmented-reality headsets to create holograms for medical training, surgery planning, and even during actual surgeries.

“This is the most dramatic and exciting time in medical history,” says Thomas Osborne, director of the National Center for Collaborative Healthcare Innovation and the Chief Medical Informatics Officer at the VA’s Palo Alto Health Care System. There’s an enormous flood of data to contend with, Osborne says, and today’s Wi-Fi and 4G LTE networks can’t keep up.

The hospital was looking to 5G to bridge the gap, and the equipment has just arrived. To be exact, Verizon’s “network on wheels” arrived on Jan. 31, Osborne says, providing a local area network with both 4G and 5G support. “On February 4, at 7:32 p.m., we generated the first 5G signal,” he says.

There are many potential uses for 5G, but for the hospital the one that really stands out is virtual reality. VA Palo Alto has the newly released HoloLens 2 augmented-reality headset from Microsoft, with wireless 5G support. Images are displayed inside the headset, and to the user, it looks as though they appear in the air around them, something like the virtual reality “holodeck” rooms on the TV show Star Trek.

Previously, high-end virtual- and augmented-reality headsets required cables, limiting their use in medical applications. In addition, traditional wireless networks don’t have the speed and bandwidth required. “We used the [HoloLens 2] system on 4G to try it out, and there’s latency and jitter. With the addition of 5G networking, we can have these 3D images [appear] as big as a room, and it was really amazing,” Osborne says. “It was exactly like the Holodeck.” 

In the three weeks since the 5G network went live, the hospital has conducted demo days for staff, who got to try out AnatomyX, a medical training application from Medivis that helps users learn about human anatomy using three-dimensional images.

The next step is to use the technology to plan surgeries, and to give doctors the ability to “see inside” their patients before they even make the first cut. A 3D holographic image would be superimposed on the patient, making it seem to the doctors as if they were viewing through patients’ skin to the organs inside, Osborne says.

Doctors would be able to see the exact location of injuries, nerves, or blood vessels, which could streamline surgeries. “And the smaller the incision, and the less time under anesthesia, the faster the recovery,” says Osborne, who is himself a doctor, having started his career as a neuroradiologist at Harvard Medical School’s Massachusetts General Hospital.

There are some hurdles to overcome, however, before the VA can use actual patient scans before and during surgeries. One hurdle is related to privacy; any new technology must be vetted to make sure it’s in compliance with data-protection regulations including the Health Insurance Portability and Accountability Act (HIPAA). “We need to get privacy and security permissions,” Osborne says. The technology is there, though, he says.

5G coverage conundrum

Outside of private deployments, 5G coverage remains an obstacle. All the major carriers, including AT&T, Verizon, Sprint, and T-Mobile, are promising 5G connectivity, but in practice it’s limited to a few areas in the biggest cities. Consumers don’t have 5G-capable phones yet, so the carriers’ 5G promises are little more than marketing hype for the time being.

Gartner, for example, places 5G at the “peak of inflated expectations” in its most recent hype cycle report and predicts that it will take two to five years before 5G reaches what the analyst firm calls the “plateau of productivity,” when mainstream adoption starts to take off.

Until that happens, many enterprises are circumventing the lack of coverage by deploying private 5G in factories, college campuses, hospitals, office buildings, or other contained environments – just as the VA Palo Alto hospital did.

“We believe that enterprise deployments have the potential to be the most significant and leading set of use cases for 5G,” says Dan Hays, principal and head of US corporate strategy practice at PricewaterhouseCoopers. “That’s unlike any prior generation of mobile technology. In the past, mobile technology has really been about the consumer, whereas 5G is really about enabling richer experiences and a larger scale of connectivity.”

But setting up 5G isn’t as easy as buying a Wi-Fi router and plugging it into the network.

For starters, the 5G spectrum is controlled. As a result, companies have three main options. They can get spectrum from the public providers like Verizon or AT&T; they can buy their own spectrum; or they can use the “general authorized access” public spectrum, part of the Citizens Band Radio Spectrum (CBRS) that has been recently approved for commercial 4G and 5G use.

Going with a big cellular service provider is easier, quicker, and comes with technical support. That’s particularly important in the early days of 5G when the technology isn’t fully matured.

Glass manufacturer Corning, for example, opted to use Verizon for its 5G spectrum, as well as for network equipment, antennas, 5G radios, and technical support. Unlike regular public cell coverage, where the data travels via Verizon’s cellular network to the public Internet, the data stays within Corning’s control.

Corning started putting together a team two years ago to plan the company’s 5G deployments. “We have to stay on the forefront of what’s happening in manufacturing,” says David Loeber, a vice president and program director for 5G at Corning.

The company considered applying for its own spectrum at auction, but the process can be complicated. “It made sense to use Verizon’s network since they were our partner in this,” Loeber says.

The first use of 5G for Corning is at its Hickory, N.C., plant, where the first 5G-enabled automated vehicles hit the floor in the second half of 2019. “We’ve got parts, materials, large spools of cable that have to move around the plant,” Loeber says. “We’ve got to get stuff from one process to another, to inventory and out of inventory.”

Previously, the vehicles could be programmed to go from one spot to some other spot, but with high-bandwidth connectivity they can receive more information that makes them smarter. “Now, the robots can move around,” he says. “They become more efficient, they become more intelligent. They’re aware not only of what they’re doing but also of the environment, of where the other robots are, and where the people are. So it helps enhance safety.”

Wi-Fi couldn’t handle this use case, Loeber says. “The WiFi spectrum gets pretty crowded with lots of different devices, and you can’t really put mission-critical devices on a network where you can’t count on getting a response back in time, and where you can’t even guarantee service.”

Today, 4G LTE can do some of this, he says. “But the number of connections is limited with LTE, and it expands with 5G by a factor of 10.”

Role of CRBS in private 4G, 5G networks

One downside to getting connectivity via a public carrier: The data charges can rack up quickly. (Corning declined to share financial details of its agreement with Verizon.)

Metered pricing can be prohibitively expensive for high-bandwidth applications such as self-driving vehicles, robots, computer vision systems, surveillance cameras. “It’s way too expensive,” says Lindsay Notwell, senior vice president of 5G strategy at Cradlepoint, a 4G and 5G networking vendor.

Since the whole point of 5G is to send more data, faster, on more devices, some enterprises may prefer to go it alone. That’s where CRBS comes in.

Late last month, the FCC approved full commercial use of the “general authorized access” tier of the CBRS spectrum. (In June, the FCC will begin auctioning off spectrum in the “priority access” tier, county by county. The third tier, “incumbent access,” is reserved for the U.S. Navy and other federal government uses and satellites.)

The general access tier of CBRS is available today, in every county in the continental U.S. and in some portions of Hawaii, says Dave Wright, president at CBRS Alliance, an industry group supporting CBRS adoption.

To use this spectrum, companies need to buy equipment, set up their networks, then register their spectrum with a spectrum access system provider. Today, those include CommScope, Federated Wireless, Google and Sony. The registration helps protect one company’s signals from interfering with others in the same immediate area. (See also: New CBRS wireless can bring private 5G to enterprises) 

“Everything that’s necessary is available today and ready to roll,” Wright says. “But we think the initial private deployments are going to be focused on private 4G LTE deployments, because there’s a much bigger ecosystem available today for equipment.”

Fortunately for early adopters, CBRS isn’t just for 5G but for 4G LTE as well. That makes the migration easier, Wright says. Companies can start with 4G LTE and then upgrade to 5G as more devices hit the market and prices start to come down.

Waiting on 5G devices and standards

Meanwhile, device makers have been working to release gear that can take advantage of 5G networks. There are around 50 certified 5G radio devices already on the market and 55 client devices, says CBRS Alliance’s Wright.

But it’s not enough to have 5G radios and antennas, says Corning’s Loeber. IoT devices, like robot arms, security cameras or Corning’s automated vehicles, may need to be reengineered to take advantage of 5G capabilities.

“Let’s say you’re trying to address automated guided vehicles,” he says. The vehicle “has to have an interface, and that interface also has to be designed and defined – it’s not just one piece that has to change.”

Multiple parts have to work together, and they’re not all in place yet. For example, Loeber says, the initial deployments of 5G are actually a hybrid of 4G and 5G. “The data is 5G, but the controls for the network are 4G because all the network control infrastructure today is 4G,” he says.

Likewise, early adopters will have to wait for 5G standards to develop further before promised attributes – including higher bandwidth, lower latency, and support for more devices – become a reality.

“The first standardization was issued just over a year ago and focused on superfast broadband speeds,” says Dan Bieler, an analyst at Forrester Research. “If you buy 5G now, you get superfast broadband, but you’re not getting other benefits that people are hoping to get from 5G, in particular the super-low latency and the mass IoT-type of scalability of connecting lots and lots of devices in a small geographic area.”

Those two features will come in the next round of release cycles, Bieler says. The 3GPP mobile telephony standards project, which represents the seven major standard-setting bodies, says that the next version of the 5G standard, Release 17, which will support low latency, will be finalized in September of 2021. Bieler says he expects the next release, with support for a much larger number is devices, to come about 18 months later.

“And then it takes time for device manufacturers to design the new 5G [cell phone] headsets and infrastructure,” Bieler adds. “And then you need to manufacture it and sell it. Normally, there’s at least nine or 12 months – or even longer – between a new standard being issued and the real availability of new equipment in the marketplace.”

Meanwhile, some providers, including Verizon, are trying to help enterprise customers deal with the latency problem by offering edge-computing resources to reduce the length of the trip that the data takes between 5G-enabled devices and corporate computing resources.

“In hospitals, you need to have very low latency if you’re building a hologram,” says Heidi Hemmer, vice president of technology at Verizon. “In manufacturing, you need to make sure there isn’t a lag between the computer sending the message and what the robotic arm is doing.” Edge computing “really brings the roundtrip latency down significantly,” she says. “We are doing trials with that currently.”

One of those early customers is Corning. “Latency comes partially from the network, and partially from the server configuration,” says Corning’s Loeber. “We have them on-premises or in a centralized office that’s not too far away.”

For the rest of the low latency and other 5G improvements, he says, “we just have to wait.”