The digital wave known as the industrial internet of things, or IIoT, is still an emerging phenomenon, but for those individuals and companies hard at work creating these connected networks, they already see a world of possibilities—one they believe will represent a historic inflection point.
“This is the fourth revolution that’s going to occur in industry,” says Tony Neal-Graves, vice president of internet of things group and general manager of the industrial and energy solutions division at Intel. “It’s going to change the way that we manufacture and deliver services to consumers. It’s fundamentally going to change the way that we do everything.”
Just as mechanization, mass production, and automation transformed industry over the last century and a half, so too will the IIoT.
Here, Neal-Graves, whose group focuses on developing IIoT solutions, discusses the shear complexity of the industrial internet and the changes it could deliver.
What’s so revolutionary about the industrial internet?
We’re going from a world where most factory environments are automated, but they still have a lot of manual intervention. The industrial internet is taking us to a world where the environments become truly autonomous—the factory makes decisions on its own.
You can instrument a machine today and collect all kinds of data about its performance and send it somewhere. You start sharing information from that machine with others in your process flow, and you can now make better decisions because of the analytics. The vision is that the machines start to optimize on their own. If equipment detects a failure, it does the most immediate thing it should do: It just shuts itself down. If a machine has the ability to interact with the data that it’s generating, if it becomes intelligent, it can take actions like that on its own.
When do you think we’ll see that kind of autonomy?
I think it’s a 10- to 15-year journey before we get there, but governments are starting to put a policy discussion in place, because everybody’s fear is “If I don’t invest in this, I’m going to get left behind.”
What does the early industrial internet already improve upon at this point?
One of the common use cases is this concept of predictive maintenance as opposed to scheduled maintenance. Companies are saying they can’t afford to experience downtime—it’s so costly—but they have to be able to take their factory line offline periodically to do maintenance to prevent an equipment failure. If they could have enough data to predict when a piece of equipment might fail, it would completely change the paradigm. That’s an application folks can rally around.
What are the biggest lessons for Intel from its work developing these new technologies?
The biggest learning that we’ve found so far has been the fact that there’s no one company out there that can stitch together a complete end-to-end IIoT solution today. It takes cooperation with some big companies, which is why we’re working with GE, but also with some small companies.
This work has also stretched us as a company. We’ve had to acquire new skills. For my organization, more than 60% of the resources that I have are people from outside of the silicon manufacturing industry.
Finally, what we’re trying to do is use our factory environment as a living lab. We tried out some predictive maintenance solutions within one of our own factories. We trialed some technology that leverages machine learning and were able to accurately predict when a pump was going to fail. It’s very compelling to see that sort of benefit firsthand.
What triggered the accelerated growth of the industrial internet over the last couple of years?
The smartphone is the genesis of this. We’ve now given everyone an “edge device” that not only communicates but also has incredible computing power. As we continue to drive down the cost of memory, computing capability, and connectivity, you can put that technology in any device. These large pieces of machinery are now capable of communicating with each other.
How do you get people outside of industry excited about the transformation that you’re expecting?
I’d say for the very self-motivated reason that companies will be able to deliver more goods and services to you at a lower cost point and give you more choice and flexibility. I talked to one major auto manufacturer who would love to be able to deliver a car to you on the same cadence that you could buy something from Amazon. You get to design your car online and then they deliver it to you. That’s where the world’s going.
There’s another level that I think about, too. If you can make industry more efficient, you can make it more green. You can use less power. You can develop products with a lot less waste.
What are the challenges to widespread adoption?
One barrier to trying new things in a factory or in an oil field is just the cost of deploying a new technology and not knowing whether or not it’s going to deliver any kind of returns. “If it works, don’t touch it” is really the mantra in the industrial world. You have to show the payoff.
What do you need for a successful IIoT conversion of a factory?
You have to be willing to sit down in that factory environment and be very transparent. To talk about what you know and what you don’t know. To ask a lot of questions to understand the problems that factory owners are facing to drive more efficiency. What you need to be successful are teams that combine software development, data analytics, and operational knowledge so you understand how the factory works, because instrumenting every piece of equipment isn’t cost effective. Instrumenting the ones that really matter is what’s important.
Right now, the industrial internet offers incredible promise and, of course, challenges. This is complicated stuff, and it’s all still coming together. We need big companies and small companies and all the smart people out there who can bring really cool ideas to the table. What’s clear is that the early innovators and the early adopters who incorporate these technologies today will be the ones who reap the greatest benefits going forward. But first, you’ve got to get on board.
This article was created and commissioned by GE and Intel, and the views expressed are their own.