The Do’s and Don’ts of Industrial IoT

Your boss wants to start an IoT project—what next?

Advertisement

In a home or on a manufacturing floor, there’s no point in connecting something just so it’s online. You must have a reason.

Andy Rhodes, the former vice president and general manager for Dell’s Internet of Things (IoT) commercial sales division, once told me that he no longer takes phone calls from companies asking for “some IoT.” If the company doesn’t know what it wants to do with connected devices, it isn’t ready for a sales talk.

So how does a management team decide what it wants from the IoT?

There are two paths here. The first is to start small with a project that can offer a short-term return on investment (ROI). In a factory, this could mean installing cameras and sensors on a manufacturing line to replace human spot-checkers. Cities might install sensor-infused LED lighting to reduce energy costs.

The second path is a bit more visionary: Build a platform or product that can change your entire business model. For example, in 2015 Emerson Climate Technologies launched a new connected product in its HVAC business. With US $100 worth of sensors per home, the HVAC installer and Emerson could guarantee a period of time that a customer’s heating or cooling system would operate before needing repairs.

The ROI on this project is still unclear. But the idea was that tracking how hard the AC unit was running, and how long it takes to cool a home, could change Emerson’s business model from offering a product (HVAC units) to delivering a service.

However, Emerson’s client would have to add sensors, connectivity, and a gateway. And that requires labor at every point, which means it doesn’t scale in the way that traditional IT projects scale.

Upal Basu, general partner at NGP Capital, is concerned that focusing on short-term ROI leaves a lot of money on the table. He says the value in IoT deployments comes from data gathered by sensors—not from some centralized platform.

But a long-term approach is far more costly to execute. Which is why most companies will try to tweak small production lines or other minor functions within their businesses first.

Once a company has decided on a problem to solve—whether it’s saving energy by turning off lights or using sensors to gather information on a factory floor—the next step is to figure out who has done a similar project and talk to them about their process.

Peter Zornio, chief technology officer at Emerson Automation Solutions, says most clients turn to his team to understand how Emerson’s automated plant operations would work for other facilities.

“Unless you have a sparky operations guy who is reading about this—they may come to you with a specific example and solution—most customers come to us because we’ve already done this,” he says.

When it comes to assessing ROI for a deployment, there are so many metrics and factors involved that it can be daunting. Even tech giants such as Microsoft and Cisco have signed partnerships within highly specialized industries like manufacturing, health care, and construction to develop the expertise they need to help clients.

Essentially, a large-scale industrial IoT project isn’t something you should ever try to do alone.

This article appears in the May 2018 print issue as “How to IoT.”

Advertisement