To find out about the impact of 5G mobile broadband service on the IoT/IIoT, I interviewed Jai Suri, Vice President, IoT and Blockchain Applications Development, Oracle, and Mike Anderson, Embedded Systems Architect, and consultant in the aerospace industry. I asked them if we are close to bringing 5G to industry or whether other applications will likely come first. According to them both, it’s complicated.

“5G was designed with IoT use cases at its core, while 4G and earlier standards were designed primarily for voice and data transfer. With 4G/LTE, fast data downloads enabled streaming content, and 5G will usher in ultra-fast, low-latency communication and the ability to handle millions of devices per square kilometer,” said Suri.

It’s important to start with a couple of clarifications. The term 5G does not refer to frequency, it simply means it’s the fifth generation of wireless mobile broadband communications (See “Understanding the Implications of 5G Cellular”, Sensor Technology, March 2020.) And it is not just one frequency, it covers three bands:

  • Low-band is from 600 – 850 MHz This is not much more than the currently available 4G.

  • Mid-band covers the range from 2.5 – 3.7 GHz.

  • High-band operates from 25 – 39 GHz. Signals in this frequency range are generally called millimeter waves and they provide the extreme performance many people think of when they think 5G.

5G in the IIoT

The Third Generation Partnership Project (3GPP) is an international standards organization that develops protocols for mobile telephony. Their Release 15 focused on 5G for consumers, whereas the current Release 16 shifts the focus from the consumer to the IIoT and other industrial markets. Their intent is over time to publish standards for developing and implementing a growing body of 5G applications.

The outstanding benefits of millimeter wave 5G are high speed, low latency, high number of connected devices, and high reliability. With a 5G millimeter wireless connection, you get the kind of performance that you would expect out of a wired link. And because of its high frequency range it is not typically susceptible to electrical interference from nearby equipment.

However, the big issues are that millimeter wave 5G can be blocked by a window and it doesn't penetrate very far. That means you need to have a lot of 5G cells to be able to take advantage of it. On a reasonably sized factory floor, you’d probably need 20 or 30 radios sitting on the ceiling over your primary work cells, according to Anderson.

Figure 2. Augmented Reality applications visually overlay 3D information on top of live video to pinpoint problems and guide technicians through the complete repair process. (Image courtesy of Oracle)

Industrial applications typically demand high reliability and 5G can certainly deliver that. As Anderson explained, “the current 4G network reliability is at 99.8%. A dropped call rate of 0.2% is considered a super-reliable, carrier-grade mobile telephone network. But the reliability requirement for industrial applications is typically six-9s (99.9999%). 5G networks can deliver that level of reliability by using cell duplication, the clever use of radio spectrum, and massive MIMO (multiple-in, multiple-out) antennas.”

In order to bring a 5G signal from a telco into a factory, an antenna that looks like a miniature cell tower would have to be set up inside of the building. To use this in an industrial application, you would need a much larger infrastructure inside of the building — you couldn't just simply penetrate the building and you’re done — it's a lot more complicated. Depending on the frequency range, you may not necessarily be able to even see the signal on the other side of the building. There are girders and all kinds of things that will absorb the signal. So, you'd need multiple antennas. Whoever is setting up the factory must make sure that those antennas have adequate coverage inside of the building and that there are no dead spots.

Prime Candidates for 5G

Digital Twins

A digital twin is a real-time dynamic representation of a physical system. For example, the aerospace industry is using digital twins on a more or less regular basis. If you’re going to put a whole lot of technology into a satellite and send it into low Earth orbit, it will likely need servicing. If there are software updates, they can first be tested on earth in a digital twin — a real-time digital counterpart of the satellite — before uploading it. For the digital twin to stay up-to-date with the real thing, a large amount of data needs to be transmitted at high speed, which is perfect for 5G.

Virtual and Augmented Reality
Figure 3. Machine learning predictions can be used to automatically drive maintenance workflows to perform predictive maintenance and avoid costly downtime. (Image courtesy of Oracle)

For virtual reality, detectable latency is unacceptable. So, the high bandwidth, high throughput, and low latency of 5G is a perfect fit. One application where virtual reality is useful, rather than just entertaining, is for simulators to train operators of systems that could be deadly if anything goes wrong. For example, aircraft, railroad, or ship operators, and most definitely operators of nuclear reactors.

“I’ve been in a training room for a ship, where you see nothing but blank walls, but when you put on the virtual reality goggles, you see all the controls that you would normally expect to see in a ship. If the simulator is also a digital twin, these virtual controls will do what they would do on an actual ship,” said Anderson.

“Another augmented reality application is enabling service technicians — and especially third-party technicians — to perform a high-quality repair job. Augmented reality could guide them through the entire repair process, by providing a prescriptive model of maintenance to pinpoint the problem and visually overlaying 3D animation on a live video feed to show the sequence of steps to be performed,” said Suri.

Predictive Maintenance

5G would also be helpful for predictive maintenance in cases where you have an application with large amounts of streaming data that you want to track and analyze.

A critical application is for very large motors or generators that could cause severe damage if they failed or could interrupt important services like power generation. Real-time monitoring for remote equipment, such as monitoring gas compressors in the oil and gas industry is another critical use case, according to Suri.

Another example is, if you are an HVAC company selling rooftop units, you want to collect data not just for maintenance, but also on its usage. “The cost of using a 4G network to collect useful data is about $3 – $5 per month, so if I have a company selling rooftop HVACs and I have 100,000 units deployed, the expense will be very significant. And processing all of that data to obtain useful information further increases the operating expense,” said Suri. “If organizations could provide their customers with insights about power consumption versus comfort level, that data could be used to sell a more efficient unit — or organizations could offer usage-based pricing where they simply charge a monthly fee based on the number of hours that it was running.”

Edge Computing

5G’s promise of low latency, high speed, and high reliability, enables enhanced edge computing to provide engineers with an expanded toolbox for designing networks.

Edge computing is beneficial in a number of ways:

  • With 5G, more computing can be done at the edge because the edge can be redefined to include “local clouds.” Previously, low-latency computation had to be done at each sensor node, while less time-sensitive computing could be assigned to an enterprise-level network or to the public cloud. But the amount of computing that could be done at sensor nodes was extremely limited due to practical limitations on physical size and power consumption.

  • With 5G connections, however, data from remote sensors can be wirelessly transmitted to ‘local clouds’, consisting of off-the-shelf computers. These clouds could perform low-latency computing, which allows for decision making to be moved closer to the source of the data. Packages of meaningful information rather than raw data could then be sent to both enterprise clouds and public clouds for less time-sensitive processing and storage.

  • In addition, organizations might want the ability to protect sensitive data from public exposure. Edge systems could process data at the source and send only anonymized and summarized insights to the cloud for further processing.

Autonomous Vehicles

“Where I think you will see 5G as a killer app, is autonomous vehicles,” said Anderson. “They have lots of sensors on them: lidar, radar, cameras. These all produce large amounts of data and require very fast response times with extremely low latency. Autonomous vehicles have to assimilate large amounts of data about their surroundings in real time and respond quickly.”

Figure 4. Cloud asset monitoring display. (Image courtesy of Oracle)

Communication between a vehicle and anything in its surroundings, including pedestrians, other cars, and traffic lights, is known as V2X. This technology enables vehicles to directly communicate with each other for crash avoidance. Also, if a car is approaching an intersection and there is a traffic light and both are connected on 5G, the car could directly communicate to the traffic light to understand its current status.

“To implement 5G for autonomous vehicles, a State Highway Department would probably let out a contract out to a telco (telecommunications operator) provider, because you'd want to leverage their cell towers and their support infrastructure,” said Anderson. “For high data rates and low latency, you’d need a minimum of 1 GHz of bandwidth. And if you really want low latency, now you're talking in the 25 to 27 GHz (millimeter range), which means you need a cell tower every block in a city. In rural areas, you could probably do about 5 kilometers between towers.”

Since current telco cell towers are spaced at about 20 miles, many new ones would have to be erected. 5G millimeter waves require roughly five times as many cells as for the normal frequency ranges. The saving grace is that the size of a 25 GHz cell tower is much smaller than the existing telco cell towers.

“The reality is that at first, the millimeter waveband will be aimed at people’s personal uses, so that infrastructure will be put in place anyway. Then excess capacity will be sold to the city or to the state for use with autonomous vehicles,” said Anderson.

Logistics

Logistics can benefit from being able to track and locate large numbers of vehicles in real time. This would be a bonanza for fleet managers. So often, tracking packages has amounted to retroactively retrieving information from a data logger. However, real-time vehicle tracking systems use GPS and cellular data to transmit information about location and direction of movement to a control station. That data can be used to troubleshoot routes or optimize logistics operations. Vehicle tracking systems can also send operational information about the vehicle, such as fuel levels, tire pressure, ignition status, and engine temperature that can be used to schedule maintenance. 5G usage in this sort of application could also be a killer app.

In Conclusion

The bottom line is that 5G will have a significant impact on both the IoT and the IIoT, but this will develop slowly, some sectors earlier and some later. Overall, the timeline for significant integration of 5G into IoT/IIoT applications is on the order of five to ten years.

This article was written by Ed Brown, Editor of Sensor Technology. For more information, contact Ed at This email address is being protected from spambots. You need JavaScript enabled to view it. or visit here  .