How to justify your first IIoT trial

Jan. 4, 2016
The 5 things that will keep your project from going off the rails before it even gets going

From 10,000 feet up, the Industrial Internet of Things (IIoT) dazzles: it’s a gleaming vista of connected machines that do the jobs they were designed to do while providing the data that can tell their owners how to drive continual improvements in efficiency, safety, productivity and profitability. But come down to earth—to the executive suite, the IT department and the plant floor—and you will mostly find questions that leave different business units struggling to chart a clear path forward or even to decide whether to take the first step at all.

Where do you start if you’re the guy charged with developing your company’s first IIoT initiative? It’s an important question. According to Dave Petrucci, analytics, engineering and IoT VP at the industrial consulting company Genpact, it’s the most important question there is, and it can be summed up in three letters: ROI.

“We’ve heard the McKinseys, GEs and Ciscos of the world talk about the high-level story, but people on the floor still have questions, starting with, ‘where’s my IIoT box?,’” Petrucci said. “This is the challenge when you start talking to the different business units. There is an expectation that this is magic. Engineers don’t like magic. CFOs like it even less.”

Beyond the magic, real barriers remain, ranging from business units locked in silos to the belief that, in the end, the IIoT is just another IT investment or, worse, another dose of hype that just won’t pay off. Moving past these obstacles requires vision and practicality, Petrucci said. Vision to create new marketplaces and leapfrog competition, and practicality to drive the step change that can result in real business outcomes.

Is your business ready?

For those who are ready to begin an IIoT trial, Petrucci lists five areas of readiness that need to be identified before a successful IIoT initiative can get underway.

Mission readiness refers to clearly identifying the problem you’re trying to solve. And it needs to be specific and concrete in order to gain buy-in and maintain momentum. “If you’re just doing this to throw analytics and technology at a general problem, that’s probably not the right reason to do it,” Petrucci said. “But if you’re trying to drive out costs or create a new revenue stream or fundamentally change the way you do business, that’s a good reason to do it; and it’s a reason that has a measurable ROI.” 


Asset readiness means understanding what needs to be connected to achieve the desired goal, then determining the optimal way to make those connections. “It’s not about throwing a sensor on everything,” Petrucci said. Instead, it’s about identifying the necessary variables to run your process and determining the best way to capture the information required to analyze and optimize them. “Everyone keeps telling you that data is cheap,” Petrucci said. “And data is cheap. But data management is probably one of the most expensive things in your organization. So how do you define what data you really need and what you don’t?”

IT/OT convergence readiness. “Like it or not, the information technology and operational technology roles are colliding, and if you look at the resources that are in those roles today, they’re not the skills or even the people that are going to be in those roles two or three years from now,” Petrucci said. “There needs to be an evolution of skill sets and people who have a broader view of how these two areas interact.

Organizational readiness means having the people and skills in place to handle the challenges of IIoT design and operations. “People have been telling you for the past few years that you need data scientists and PhDs in machine learning. But real operational companies don’t have these skills,” Petrucci said. “You really need to assess if you have what’s needed to move ahead.” 

Execution readiness refers to one simple question: Can this be run in a daily operation? “There’s a difference between science projects and programs that can gain the trust of upper management to run the real operations of the business,” Petrucci said. “You need to have sufficient organizational alignment and clarity of mission to move a project from a science project to an investment that be leveraged to improve the business.”

If these five areas of readiness are less than ready, then fundamental work needs to happen before any IIoT initiative can be put into motion, Petrucci warned. But if an enterprise can confidently check off these five boxes, then it’s time to move to a pilot program.

Starting on the wrong foot

The traditional—and wrong—way to implement new technology is pick a cool technology, figure out how to implement it, then hope it will solve a business problem down the road. “If this what you’re doing, stop,” Petrucci said. The first step is to define the business problem you want to solve. “It’s as simple as going to your corporate metrics and asking, what are the three things that are going to make the CTO or CFO or CEO successful? Dig into those and define how you’re going to support those through all of this new technology.” Once these metrics have been identified, there are four steps to designing your pilot: Define the business issue you want to affect; identify the data do you need to affect it; determine how you are going to collect, store and analyze that data; determine what you need to sensor.

Remember, there is no magic IIoT box. And there is no magic path to the promise of IIoT functionality. Successful pilot programs require alignment of goals and people and multiple levels of readiness. Without these, Petrucci says no IIoT program can succeed. But with them, the promise of business transformation can be made real.