Why-5G-AI-and-IoT-combo-will-be-so-powerful-for-data-analytics

Combining 5G, AI, and IoT will help companies collect more data, analyze it more quickly, make more accurate predictions, and rapidly take action.

Thanks to its high bandwidth, low latency, and the ability to support higher end-point densities, 5G will allow companies to collect more data, analyze the data more quickly, make better predictions, and rapidly take action. I spoke with Bill Menezes, Sr. Principal Analyst with Gartner, about how 5G will affect data analytics and IoT technology. The following is an edited transcript of our conversation.

Bill Detwiler: So Bill, tell me a little bit about how you see 5G affecting IoT technology, maybe, in the near term. Is it going to have much effect at all as we look at maybe one, two years into the future?

Bill Menezes: Well, our expectation is that, even though IoT eventually will be a huge part of what 5G is going to create benefits for, in the next several years, it’s going to be relatively limited. There are a couple of reasons for that, and one is that a lot of the IoT specific technology innovations that’ll be incorporated in 5G won’t occur until the next couple of releases. And so, given those are coming out in the next three to five years, and of course there’s a lag for commercial products, IoT will be a big part of it at that point, but not really in the initial stages.

So our expectation is that, in the initial stages, the IoT that’s going to be most effected are the types of applications and implementations that require really fast data throughput, right? That’s the initial benefit of 5G. And so, we’re looking at a forecast that says the two biggest use cases for IoT, say by 2023, are going to be the automotive industry for embedded connected car modules, we think that’s going to account for close to 40 percent of 5G modules installed base, by 2023, and the second is, believe it or not, outdoor surveillance cameras, right? In government and physical security sectors. That’s going to be about a third of the installed 5G modules in that same timeframe.

So right there, you’re looking at over 70% are going to be accounted by two applications that benefit from really high data speeds and a lot of data throughput and, to a certain extent, an improvement in latency. The IoT specific elements that rely on things like massive machine-type communications and ultra-low latency high reliability are going to unfold in a longer term than that.

Bill Detwiler: And you mentioned data, and one of the reasons that throughput is so important these days is around the analytics that is being done with that data and how that data is being used beyond just the need to transmit information…in a way that allows you to process the information and then take an action on that. Talk to us a little bit about that data in terms of analytics and beyond, in the long term, once we get to those applications that are really collecting so much data. How is 5G going to play in to sort of allowing companies to expand analytics?
Bill Menezes: Sure. If you look at what we think some of the key IoT technologies, even aside from 5G are, one of the big ones is artificial intelligence, and artificial intelligence is going to be a big element in terms of how you apply that to the massive amounts of data that this huge, dense population of 5G endpoints will gather for things like predictive analytics, right?

So a combination of being able to support a really high population of IoT endpoints, in a given implementation, in a small area, for example, like a smart building or an area of a small city or an enterprise setting like manufacturing, is that applying AI to the data that comes through in much faster timeframes because of the low latency is going to enable a higher and higher level of artificial intelligence to do things like predictive analytics so that, for example, in an IoT implementation, that might be a control implementation, right? Precision manufacturing.

You are able to have the intelligence to do things at a much more granular, precise level sooner because of the ability to do much better predictive analytics that 5G is going to support versus, say, current technologies like 4G, right? So, if you’ve got a wired, fiber broadband connection, you’re able to do a lot of this stuff now. The advantage of 5G is you’re going to be able to do more of that in a setting where you don’t have to have that tethered connection.

Bill Detwiler: That was actually going to be my next question, which is how much of a game changer is it to be able to do that wirelessly? To be able to have a fiber, or better than fiber … Current fiber speeds for that last mile, for that … But also be able to have more devices. You get more devices, you get higher bandwidth and you don’t have the expense, you don’t have the time constraints of having to lay fiber, of having to run cables.

Bill Menezes: We already see some of those benefits in 4G networking when it’s used in settings like manufacturing. The ability, over time, to evolve that to a 5G deployment means you can support more of those end points and you can collect a lot more data and have that much higher level of determinism because of the low latency. Now, where we see the benefits of that are in areas … We mentioned manufacturing, right? So, if you have a facility that needs to retool every couple of years, then the idea that you’re going to be able to move around those pieces of machinery without having to rerun cable in order to get that same type of broadband connectivity to all your sensors and control end points on that machinery, that’s a huge benefit.

You can already do that with 4G, but as those end points are enabling much more sophisticated operations in that machinery, the availability of 5G and the fact that you’re going to be able to operate it with a pretty high level of quality of service versus, say, an alternative like Wi-Fi, that’s where you’re going to start seeing those benefits. And see, those are known use cases, right now though, when you think about 5G, the expectation is they’re going to be new use cases that emerge that we really haven’t even thought of yet.

If you think about the dawn of 4G, I don’t think there were a lot of people speculating that, okay, now that we’re going to have LTE everywhere, we’re going to get Uber and social media is going to be pervasive, right? With high-resolution video. Those are things that came up as it became clear, especially as developers were able to really work with the networks, what was going to be possible, and we fully expect that some new implementations in IoT will emerge that really maybe even aren’t on the radar yet.

So yeah, autonomous vehicles. Yeah, high-resolution medical imaging. Maybe you’re in to assist using virtual reality for remote surgery. Those are all possibilities, but we think that there probably will be some things that crop up that the light bulb goes on, it’s like, yeah, we should have expected that.