Back To Top
Digital Transformation

Digital Transformation: Top Down Or Bottom Up?

March 5, 2021

By Peter Zornio, Chief Technology Officer, Emerson

Digital transformation (DX for short) is a very hot topic right now. We can broadly define it as the practice of applying new digital technologies to deliver business improvements, such as transforming or automating work processes with technology like robots and mobile-connected devices; using analytics to drive better decisions; or using artificial intelligence to replace human interactions. DX can benefit virtually any business process — from hiring practices to customer and market research to planning and scheduling decisions.

In industrial manufacturing, the biggest case I see for DX is improving the core business operation: the actual manufacturing process. Many manufacturers have DX programs focused on plant production metrics such as safety, product profitability, quality, plant reliability and sustainability. Given that the revenue for these companies comes from what they produce, it makes perfect sense that this would be their focus.

And while the purpose of DX is simple, there are vexing questions that can freeze an organization. Where do I start? How do I know if I’m succeeding? Who should be responsible? Here are two common approaches I see as the CTO of a company that offers technology solutions for industrial manufacturing companies.

Top-Down Approach

Because we’re talking about networked digital technologies, companies often ask corporate IT departments to take the lead on DX programs, sometimes with help from corporate engineering or research and development (R&D). This team will then look for and trial broadly applicable solutions or architectures. The expectation is that true business results will come from the functions that utilize the selected technologies.

Companies often pursue broad initiatives such as cloud technologies, analytics or standardized deployment of other digital infrastructure while ensuring they follow companywide cybersecurity practices. In the last several years, a number of “enterprise analytics” or “enterprise IoT” software platforms have also appeared to serve this purpose. Leaders may have a broad idea of the applications they'll deploy, but they don't always know the specific cases they'll pursue. This is what I call a top-down approach.

Bottom-Up Approach

But the company “front line” is the operating unit — the manufacturing plant. This is where the action really happens, whether it's in a factory, plant, mill, oil field or mine. These facilities are typically the corporation’s source of value and have their own management, budgets and priorities. To meet targets, plant operation leaders tasked with digital transformation projects typically make decisions about investments and resource allocation based strictly on the site’s goals, such as personnel safety, equipment health, production uptime, productivity and cost containment. In my experience, no one knows better if a new work process or production change is practical and what the actual business impact will be than the operations folks at these facilities. They also know what immediate problems they would like to solve and start DX projects of their own to solve them. Let’s call this the bottom-up approach.

A Blended Approach to DX

So which approach yields the best results? As I’m sure you’ve guessed by now, I've found that the answer is a combination of both.

Facility-level folks can move fast. They can quantify the return on investment of a DX project easily because they know the business impact and how to tie in the facility-specific digital systems. They also have the domain expertise to apply to manufacturing problems — they know in detail how their equipment and processes work — something data scientists and DX technology gurus don’t necessarily have. As you introduce new digital technologies, it’s important to ensure that facility operators are willing to change and adopt the new digital tools or work practices the DX project has put in place. You can have all the greatest technologies at your fingertips, but nothing will happen unless the facility folks execute the “new digital way.”

However, at the facility level, team members are not likely to consider how to reproduce a successful in-plant digital program across a fleet or find enterprise-level uses for their data — and they almost certainly don’t have the funding to do so. Furthermore, they are often not as comfortable with advanced technologies, such as those that can help them move data outside of their own facility to the cloud. The big risk is that they create a set of facility-specific solutions with high support costs that the rest of the organization can't leverage.

Corporate-led programs will typically pay great attention to the broad applicability and standardization of an architecture or solution. Corporate IT organizations are also generally very good at governance and making data widely available and reusable. But sometimes they don’t know the specific data types and connections they need for a unique application, which can create “data lakes” that are difficult to use. They lack the needed domain expertise, so they may miss critical functionality or deliver applications that don’t deliver results. And if they don’t work hard to gain the buy-in of facility folks on their projects, they run the risk of it all being for naught if no one will execute or “close the loop."

In my experience, digital transformation only works if there is a tight connection between corporate digital teams driving the vision and field-level plant operations targeting specific, measurable problems to solve. One of the best ways to accomplish this is to establish a DX team tasked with bringing corporate- and plant-level IT, operational technology (OT), engineering and business leaders together to prioritize problems they want to solve and then implement pilot projects. It’s important to find “champions” in facilities who will lead these projects and share their experiences with other facilities. The trick is figuring out how to scale those solutions across the enterprise and avoid “pilot purgatory.” If you don't, both groups might have illusions of progress when in reality, they're just creating more silos of limited-deployment solutions. That's why you should combine the business and operations knowledge of plant personnel with the technological and support savvy of central IT and engineering to create a winning team for digital transformation.