The principal job for information technology (IT) staff is to provide the infrastructure to enable us to work: networks, printers, servers and business applications, gateways, firewalls, and now access to cloud applications — although cloud is a sensitive area for IT, since it may replace much of IT's core business.

 

Now I hear the industrial internet of things (IIoT) and digitalization present a whole new set of IT opportunities. Many times in the past working with large petrochemical companies, I have heard the IT department declare with great excitement the next wave of IT business investments that must be made. Maybe you remember this: “We need a data warehouse; we need to be doing business intelligence!” There were proclamations of technologies to be deployed, missing any concern about the absolute business problem to be solved.

 

A great technology that is a must? My response was always, “That’s the answer, but what’s the question?”

 

Today, I respond the same way about cloud, digitalization and IIoT. That’s not because I do not believe there are very substantial opportunities therein, but there’s a danger in an outlay of millions and millions of dollars on a “platform” without a clear understanding of the business problem to be solved. The platform hype is deafening. Let’s build a special mousetrap, then all we need is to find special mice!

 

 

What Are We Solving?

 

In manufacturing, we have been embroiled in digital since the introduction of the distributed control system in 1975. The cloud is “a computer somewhere else,” and IIoT is all about collecting data for analysis to uncover actionable information. It seems we have seen them all before. So what’s different? Frankly put, it’s more and cheaper: sensors, sensor networks, smart machines, smarter analytics and larger, more diverse data combinations.

 

But let’s not forget the business in our zeal to deploy the next big thing. The real question is, “What will we do differently to extract new value from cloud, digitalization and IIoT?” Asking that question will lead to the business issues and opportunities that will save costs and uncover more revenue.

 

Don’t do cloud platforms just because you can, but rather do them because it provides a way to gather data from multiple, disparate sources that can expose a new and different business prospect. The same applies to an emerging concept: edge computing. Ask, what is to be solved, and why should it be done at the edge? And what advantages does the edge bring?

 

 

The Big Opportunity in Manufacturing

 

Let’s consider an example. Currently, the No. 1 issue causing profit losses in manufacturing is equipment breakdowns. The National Association of Manufacturers suggests worldwide manufacturing is a $14T business and that 10 percent is lost to breakdowns. That’s a $1.4T opportunity that has not gone unrecognized by old and new vendors proposing solutions. Consequently, the big push is to reduce and eliminate unplanned maintenance, thereby improving equipment availability and net product output.

 

Leading-edge machine learning algorithms, neural networks and big data are deployed to exercise complex analytics to provide early warnings of failure patterns so that decisive action can fix small service and repair problems before they become expensive breakdowns. Why should this be solved in the cloud? For a single manufacturer, there’s no compelling reason. After all, the computational power can be located within the site, inside the firewall, thereby reducing potential data security concerns.

 

So why do it in the cloud? The cloud provides the capability to bring the data together from multiple sites for a single company, allowing the amalgamation of data streams from multiple machines at the same time. Such data aggregation can offer new insights across many machines, assuring higher levels of protection across pools and fleets of equipment. As financial institutions discovered, sharing credit card fraud information across companies ultimately brought benefits to all, especially the larger companies. My prediction is that the conservative manufacturing industry will eventually come to the same conclusion about sharing machine degradation and failure data.

 

Of course, there are other operations and supply chain management opportunities that can be addressed with digitalization/IIoT technologies, in local or cloud solutions, but it takes the appropriate levels of questioning to expose where and how the solutions should reside and operate.

 

So, instead of jumping headlong into building a multimillion-dollar data lake to gather data streams in the cloud, ask penetrating questions about the problem to be solved, and then arrange the data the correct way to solve it. Make the platform as big as it needs to be — no bigger — and place it where it needs to be, but ensure it is extensible when, and only when, you uncover more profit opportunities.

 

Instead of “That’s the answer, but what’s the question?” remember to ask the right questions to get to the answer.