In this episode of the McKinsey Podcast, Diane Brady speaks with partners Michael Chui and Enno de Boer about the fifth generation of wireless technologies and how artificial intelligence, the Internet of Things, and other advanced technologies are reshaping businesses. An edited transcript of their conversation follows.
Diane Brady: Hello, and welcome to the McKinsey Podcast. I’m Diane Brady. Today we’re talking about the fuel that is powering the Fourth Industrial Revolution. It is, of course, the much anticipated 5G, or fifth-generation wireless technology. What it delivers is an astonishing level of connectivity that will transform every industry. Today we’re focusing on the impact it’s already having on manufacturing, from the supply chain to how we run our factories, and, more importantly, that there are many times where we don’t actually need 5G to get this done.
Joining me today are two McKinsey leaders who spend a lot of time on the front lines of transformative technologies. Michael Chui is a San Francisco–based partner at the McKinsey Global Institute, where he leads research on the impact of technology on business and society. Enno de Boer is the global head of manufacturing out of New York, where he’s worked with partners like the World Economic Forum on the future of production and the Fourth Industrial Revolution. Enno, Michael, welcome.
Michael Chui: Thanks, Diane. Great to be here.
Enno de Boer: It’s great to be here.
Diane Brady: You are both on the front lines. I want to know who has the cooler job here. Enno, what is your answer on that?
Enno de Boer: I thought it was Michael, but my job got more and more exciting over the last three, four years—as technology is just exploding on the shop floor. So I would say I’m now head-to-head with Michael.
Diane Brady: Well, then we have to start with Michael. Just give us a primer on connectivity, because there really are two things I remember about the Super Bowl last year: one is the halftime show with Shakira and J Lo, and the other is this slew of 5G commercials announcing that the future was here. So is it?
Michael Chui: Well, it’s on its way here. You know that the G in 5G stands for generation. This is the fifth generation of these wireless technologies. And each one has been faster—and that’s definitely true. But there are other things, other benefits that this generation of technologies brings. Number one is, it actually reduces latency, which is this slightly technical idea.
But it just means that signals go one place and back much more rapidly. And that’s really important when you need to be highly responsive, like for a self-driving car, for instance. It also provides the capability of the Internet of Things, including things on the manufacturing shop floor, which I’m sure we’ll discuss later, and reducing the amount of battery necessary for something that isn’t powered up to stay connected. So it does a whole bunch of different things. The other thing that’s really valuable is it’s much more efficient in terms of how it uses radio spectrum.
Diane Brady: Enno, is the fifth generation what’s powering this Fourth Industrial Revolution?
Enno de Boer: It is, and it is not yet. But it is soon to come. So to your earlier question, I think 5G is just coming on the shop floor because you need private networks, private 5G networks, and those are just coming up and becoming available. Now, we have a real connectivity challenge on the shop floor because typically to drive value off the shop floor you need to bring together 20 to 30 digital solutions that are really changing the way you run your operations and really augment the operator.
I think the challenge that manufacturers are seeing now is, as we see this suite of technologies coming toward us, how to deploy them in a very structured way and make them work together. So that’s the challenge at the moment—it’s all about deployment. The technology is there.
Diane Brady: I want to start with COVID-19, because we’ve seen a number of both unanticipated and, to some extent, anticipated fallout from that. Michael, what has it done in terms of connectivity IoT [Internet of Things]?
Michael Chui: A couple things. One is that it has brought to the forefront some of the applications that advanced connectivity—5G and other technologies, actually—can enable. So if you think about things like trying to understand whether or not your workforce is safe, and being able to determine where they are, who has been connected to whom, if you need to do contact tracing, and those sorts of things, these technologies enable those types of applications.
It also enables remote healthcare. A number of different things, which could respond to the real public-health and healthcare needs that this unfortunate and terrible pandemic has caused. At the same time, as Enno was saying, we are just in the process of installing all of this 5G infrastructure. It will take billions of dollars and countless person hours to actually put all of the base stations, and antennae, and all the other things that you need to put in place.
In fact, it has slowed the deployment of these technologies because it’s harder to operate. These are people who have to work in the physical world, the people who you might describe as essential workers. They are falling sick, they need to maintain social distance, they need to have PPE [personal protective equipment]. And, capital budgets have been strained. The industry association estimates that deployments for 2020 are down about 25 percent as a result of COVID-19. So the need has been increased to a certain extent, or become more obvious. And yet the challenge about deployment has also increased, particularly while we’re still in this pandemic.
Diane Brady: Enno, my interpretation, or at least what I’ve gleaned from this pandemic, is that it has accelerated digitization. So I’m trying to bring those two ideas in line—that there’s less money, there’s less being done. But yet, at the same time, we’re clearly seeing acceleration on some fronts.
Enno de Boer: So we see on the shop floor, in the supply chain, an absolute acceleration. And the reason for this is that digital is driving real business outcomes. We have seen in this pandemic that we have a lot of uncertainty in the system, uncertainty that needs us to be very closely connected to the customer, because connectivity is important to the customer. And we have seen a lot of supply disruption in the system through this pandemic.
Supply-chain resilience is now a must-have. And if you think about supply-chain resilience, what you need is to connect literally all the way to your supplier and really understand where they are in the production process. You need to have a control in place that gives you oversight over your facilities. And that needs a lot of connectivity.
We have seen a lot of supply disruption in the system through this pandemic. Supply-chain resilience is now a must-have.
So what I have seen in this pandemic is that the front-runners in the Fourth Industrial Revolution have done very, very well because they were prepared and they could retool their systems very quickly and react. Everyone has acknowledged the need for digital technology on the shop floor, so the race has very much accelerated.
Diane Brady: It reminds me of that Warren Buffett quote, “When the waves go out, you see who’s been swimming naked.” And that is almost the situation here—those who were prepared have done better. And so, Michael, is it accelerating across the board?
Michael Chui: No, it isn’t. And as we often find in this digital realm, it does tend to increase the dispersion, the performance difference between those who are leading and those who are lagging. One of the other reasons why acceleration has been able to occur is, when we studied hundreds of different potential use cases, different ways in which you can use advanced connectivity, about 70 percent to 80 percent of the value could actually be achieved by existing technologies.
Now, adding 5G and other technologies which are being rolled out will increase the performance even further. And yet, for those who have already invested in having the infrastructure in place, and are able to put some of these other enabling technologies in place, they’ve been able to accelerate. And clearly the demand’s been there, as Enno mentioned. Whether it’s certain types of retail volume in grocery accelerating because people are staying at home, there are places in the economy where demand has actually increased.
But even if your demand hasn’t increased and you do have this uncertainty, you’re going to have to manage your supply chain, you’re going to have to manage your workforce. And so we are starting to see all of these things happen. The more digital a sector gets, the bigger the difference in performance between the leaders and the laggards.
Diane Brady: So it’s like a big day of reckoning in a way. I want to hear more about how this is actually playing out, Enno. Tell me a little bit first about this Global Lighthouse Network—what is it?
Enno de Boer: Yes, we started three years ago, with the World Economic Forum, to look at the top 1,000 manufacturers and how they are adopting technology on the shop floor and throughout the entire supply chain. And we have looked for examples of lighthouses that are adopting this with impact at scale.
Diane Brady: So that’s like a Michelin star for connectivity? Is it a designation they get for being further along than their peers?
Enno de Boer: I would say yes and no. A Michelin star, yes, because these 54 lighthouses—and we just announced the next ten—are really the top front-runners in the Fourth Industrial Revolution. In that sense, I would say, a Michelin star. But the ambition of the network is really to make sure that everyone learns from these front-runners and everyone is taken along on this journey. And so we are hoping that everyone is coming at some point to this standard so that this doesn’t stay a Michelin star.
Diane Brady: It’s interesting that, to Michael’s point, not all of this is the latest and greatest technology. Some of it is connecting the stuff we already have. Is that true for the lighthouse?
Enno de Boer: It’s very true. Let me just give you a fact. You would be very surprised, but today only 20 percent, 30 percent of all factories around the globe—we have ten million—are really connected with Wi-Fi.
Diane Brady: Really? There’s no Wi-Fi in the factory?
Enno de Boer: There’s no Wi-Fi in the factory.
Diane Brady: I didn’t know that.
Enno de Boer: It’s very interesting. Very basic technologies are not there, so this all needs to be put in place. If you are a company that has a factory that still does not have Wi-Fi, you might not need 5G right away. But we’re getting more and more [factories] that are getting there, running 20, 30 use cases that are fundamentally driving value, driving agility, sustainability.
We heard from Schneider Electric. They connected their entire energy-consumption sources and reduced their CO₂ emissions by 78 percent through this technology. If you want to do that, you need to have a different kind of technology.
Michael Chui: Just to be fair to our manufacturing colleagues, a lot of factories are not very radio-friendly with regard to Wi-Fi. It can be a little bit harder than setting up Wi-Fi in your house, which, by the way, can be challenging as well. But nevertheless, Enno’s point is a great one. There’s a lot more to be done in just very basic connectivity in a lot of factories.
Diane Brady: I know you talk a lot about artificial intelligence, Michael, that it may be even further off, that we may use that term a little too loosely at this point.
Michael Chui: Well, I think there are lots of reasons why we might use that term a little loosely. I think it [artificial intelligence] implies C-3PO’s going to come in and run your factory. But at the same time, I would point out the following—and we’ve noted this a lot as we’ve thought about artificial intelligence and Internet of Things—these sensors and devices that are connected in a factory, as Enno knows very well, many times people think, “Oh gosh, I have to install all these sensors and install this infrastructure.”
I think AI implies that C-3PO’s going to come in and run your factory. But at the same time, you can use artificial intelligence and analytics oftentimes with data which is already being generated.
In many cases, these machines are already throwing off data—people have used this term “exhaust data” or “data exhaust.” The machine just throws off this data. You can start to use artificial intelligence and analytics oftentimes with data which is already being generated. And so I wouldn’t necessarily say you have to have everything wired up before you can start using AI. In fact, you can start using AI, which, by the way, is to a certain extent just a bunch of statistics. Don’t think about it as being Star Wars. You can start using that tomorrow, and we see lots of companies that already have started using the data exhaust, which they just happen to have sitting around.
Enno de Boer: And the manufacturing sector has more data than any other sector already. So at the moment, the only challenge is that we only take less than 1 percent of the data to make a decision. So I’m there with Michael. There’s a lot that can be done with AI. It’s already creating a lot of value on the shop floor.
Diane Brady: What are some of the interesting use cases?
Enno de Boer: Number one is the Schneider Electric case, for example, where they had sensors on every energy-consumption [source]. And they would take this in, they would predict the consumption, and with that, optimize the consumption. That would not only reduce their energy consumption, but also their CO₂ footprint. So yes, I have three examples for you.
The first one is, it’s all about prediction. So for example, energy consumption. You can predict energy consumption and then optimize this. Schneider Electric talked about how they had connected all their energy sources via IoT, and then were predicting and optimizing the consumption that reduced their CO₂ emission by 78 percent. Another good example is to predict failures in machines and, with that, prevent the failures before they happen, keeping the machine uptime higher.
Another one is using forecasting to forecast demand. It’s a very common use case where you really try to better understand what demand is coming, and prepare your network so that you can supply that demand.
Diane Brady: If I’m sitting in the middle of a pandemic, and I am worried about my expenses, I can see the return on my investment, but I don’t have a lot to invest at the moment. What is the advice that you give to those that want to get some of the gains to this but can’t afford to retrofit their whole factories? Where do people tend to start?
Enno de Boer: Yes, so I think we have studied this very well, and we have seen a lot of companies and organizations wasting a lot of resources and money because they are following the pilot purgatory. They are following the approach of letting 1,000 flowers bloom. They are doing a lot of POCs everywhere around their network.
Diane Brady: What is a POC?
Michael Chui: A POC is a proof of concept. It’s literally where you’re testing out whether the technology works. Now, the problem is that a lot of organizations are doing a lot of these POCs because they’re a little bit in doubt, and they’re trying to figure out what to do. Now, we are saying, “Look, they have been tested. Go to the places, to the lighthouses, for example, that have tested them and gotten the value out of it.”
The approach that we have seen working in the lighthouses is, figure out one place—for example, one factory—and concentrate on that. Figure out what is the business problem you want to solve there. Is it about agility? Is it about productivity? Is it about sustainability? Is it about resilience?
And then compose the right set of use cases and make a dramatic change, and while you’re doing that, build a scale-up vehicle that allows you to take this across your organization. That is a very wise way to spend the scarce resources that you have on the digital journey.
Diane Brady: Michael, I think right now about the concerns people have around hygiene, social distancing. And I would want to have as few people as possible in my plants, which then raises the specter of job loss and automation. How’s that playing out?
Michael Chui: The question about automation and its impact on occupations and the future of work is another area of study at the McKinsey Global Institute. We do think, and we have observed over time, that more and more activities that were previously done by people and with human hands are now done by machines. But in history, what we’ve discovered is that we’ve found new activities for people to do. And so one of the questions we try to ask in our research is, “Is that changing?” because we have so many more technologies now.
What we actually discovered is there’s still plenty of work that people will need to do going forward that is hard for machines to do. As a result, one thing that we’ve been saying is, the greatest challenge going forward is probably not going to be mass unemployment because of automation. The requirement for mass redeployment from what people were doing previously—perhaps doing things with their hands, and then being able to manage a whole bunch of robots, or move into other fields—is a huge challenge that I’ve described as a great, grand challenge for the next few decades.
What we actually discovered is there’s still plenty of work that people will need to do going forward that is hard for machines to do.
But it’s a very different challenge from the idea that everyone will be out of work because everything will be done by machines. We don’t see that coming in the foreseeable future. Now, by the way, there is another challenge around unemployment that’s related to the COVID-19-related recession. But that’s a different set of challenges. It’s not simply because all these robots are going to take everyone’s jobs.
Diane Brady: More of a skill challenge, you mean?
Michael Chui: Indeed.
Diane Brady: Enno, I want to get this back to 5G. How important is the 5G or the advanced-connectivity portion of this to success?
Enno de Boer: Yes, I would say there are two answers to that. Number one is that there’s certain use cases where you really need 5G. So, for example, if you want to run automated guided vehicles, AGVs, over the shop floor, they need to communicate with very low latency, not get in traffic jams, and really get to the line. For that you need the kind of technology that 5G is offering.
Another example is if you want to use augmented reality to instruct experts that are bringing their expertise to the line, but then have even more experienced people in the background who, with augmented reality, give them clues on how to do their jobs. That is also something where you need very low latency, you need good bandwidth, and 5G technology is really helping to do this.
There are a few of these use cases that can only be done with 5G. But then I think if we’re stepping back, as we’re seeing an explosion of the shop floor use cases, we found 92 high-impact use cases in our research with the World Economic Forum. If you bring them at the same time to the shop floor, you’re running into bandwidth problems. The great thing about 5G is that you can prioritize the traffic like a traffic cop, and with that, you can then manage how the bandwidth is flowing, and you can make sure that literally every application is running properly.
Diane Brady: I hear a lot about the speed of disruption. So since we’re now talking about the acceleration—the explosion—it would suggest that the timeline for action is pretty quick. If you’re somebody that has not done this, how quickly will you be left behind in the dust in some industries?
Enno de Boer: Let me answer that for manufacturing. So we have these 54 lighthouses, and I would say, compared with their industry peers, they now have an advantage of roughly 18 to 24 months, compared with someone who has not figured out how to leave the pilot purgatory. In technology terms, that is a scary long time.
Now, do we see already the effects? Do we see that they take their entire industries, their entire markets? No, we don’t see that. But if we spin this forward, and if we are not getting the masses really going and getting on the same kind of flywheel as these lighthouses are, then we have a real problem. Because I think then a lot of companies and organizations will be left behind, and then in every industry you will see a few that have really figured it out, and they are creating an immense competitive advantage.
Diane Brady: So, Michael, there’s a lot of noise out there, whether it’s certain theories about 5G or who’s ahead, who’s behind. How important is it that we resolve the societal discussion around this?
Michael Chui: Well, different things are of different importance. At the broadest level, you can talk about the societal importance of this, particularly in advanced economies in the West or Japan, but even China. Because of demographics, the sizes of our workforces are decreasing. And what that means is, in order for us to continue to have economic growth so that we can take people as they retire, et cetera, we actually have to increase productivity.
We need to increase the amount that every worker produces for every hour that they work. Because we’re just simply going to have fewer workers. These advanced technologies—these advanced communications technologies, artificial intelligence, Internet of Things, the things that Enno is talking about and helps his clients do in the factory—can help accelerate productivity just so that we can make sure our children have better lives than we do. It’s hard to think of something that’s more important than that.
And so, getting these technologies in place and putting the management innovations in place that allow us to use those technologies better is incredibly important, you know, for our children, just to be cliché about it. At the same time, this means that people have to change what they do. As the labor force changes, we need to also pay attention and make sure that people actually get paid for the work they do, because some of these technologies could potentially increase inequality.
That is something that I think we need to pay attention to. And quite frankly, I think these skill shifts and income changes are the things that we need to pay attention to much more than worrying about how everyone’s going to be out of a job because the robots will take everyone’s jobs. It’s more like, “How do we actually keep people working and make sure that they’re paid, respecting the value that everyone produces for the economy?”
It’s one of the reckonings of COVID-19, right? We discover who is it that we actually consider essential workers, and how much do we pay them? I think that’s a really important set of things, and it’s something that this pandemic has really laid bare.
Diane Brady: Michael, help us see a little bit around the corner as to what’s next in this area. What’s on your radar?
Michael Chui: One is I think we’re going to continue to monitor what the cutting edge of technology is. Things continue to change—on the connectivity side, on the artificial intelligence analytics side, in the machines themselves on the shop floor. So we’re going to continue to monitor that. But perhaps more important is, how quickly can we put these technologies in place and pair them with the management innovations that will actually increase productivity? How can we see competition drive companies forward? Those are some of the things that will really determine how things play out, particularly as we’re trying to recover from this pandemic-related recession.
Diane Brady: What about you, Enno? More lighthouses to come?
Enno de Boer: Yes, more lighthouses to come. But more important, I think, is that the digital natives have also come into this space. The latest lighthouse that joined our network was from Alibaba, which totally innovated the way apparel manufacturing is done. With customer insights connected all the way to the shop floor, they are able to reduce their delivery times by over 70 percent, their product-development time for new apparel by 75 percent.
I think that will show a lot of industries that it’s time to move and will even further accelerate what we have seen over the last couple of months. It’s really accelerating. So that will innovate an entire industry. And I think all industries should be paying attention now. Once a digital native’s coming in, and creating their lighthouses, I think then it’s too late [to catch up]. So now is the time to get seriously moving on the digital journey.
Diane Brady: And any final thoughts? I’ll start with you, Michael.
Michael Chui: Well, one is, break through the hype on 5G. Really understand what that technology is. It’s not just a way to download things faster. In fact, it has a whole bunch of different characteristics that you need to understand. And particularly, as Enno mentioned, deploying your own 5G network within a factory, I think that’s really important. Secondly is, understand the use cases. Enno mentioned a few of them, which really can create value for your business. And thirdly, how do you scale? How do you actually get from pilot purgatory to impact that your CFO can talk about at the next earnings call? I think that’s one of the biggest challenges for an organization, for a company.
Diane Brady: Great. Enno?
Enno de Boer: Well, I would just say don’t wait. It’s not too late.
Diane Brady: Michael, Enno, thank you for your time.
Michael Chui: Thanks for having me.
Enno de Boer: Thank you.
Diane Brady: That was Michael Chui and Enno de Boer. If you’d like more information on their research, including more information on the Global Lighthouse Network that Enno alluded to, go to McKinsey.com. Thank you for joining us. I’m Diane Brady.