Get ready for a lot of talk about digital twins

When it comes to industrial IoT, it’s pretty clear that sustainability use cases are driving more investments, although a new report from Microsoft shows how ROI is still a bigger use case in industrial automation. The report, which was released earlier this month, also shows how far companies have come in adopting smart manufacturing, and anticipates that the next investments will be in digital twins, building connected products, and edge computing.

Microsoft, Intel, and IoT Analytics compiled the report from interviews with more than 500 executives at companies implementing smart manufacturing. It provides a snapshot of where we are as we emerge from the pandemic and what those investing in industrial IoT are looking for. Let’s get into it.

Four or five years ago, the big talking point in industrial IoT was “pilot purgatory,” which described a number of efforts by companies to deploy sensors, connectivity, and AI to aspects of their manufacturing operations that would get bogged down in testing. It would leave companies stuck in pilot purgatory because they had a hard time figuring out how to scale beyond the initial use case.

Luckily, the manufacturing industry has since moved beyond that with 72% of those surveyed already implementing a smart factory strategy and almost two-thirds (65%) in various stages of implementing their overall IoT strategy. Not only is this great news, but it begs new questions, namely what sorts of tools are manufacturers eying for their IoT? And what are the use cases they want to pursue?

Connected products, edge computing, and digital twins are the next elements of a smart factory strategy the survey respondents indicated they wanted to pursue. The most popular connected products are monitoring dashboards to track equipment, inventory management software (to avoid running out of necessary materials), and location tracking.

So far, a third of the respondents’ revenue is generated from those three connected products. They moreover expect that, in the next three years, the number of smart products that might report into such dashboards or get tracked by inventory management systems will rise to 47%.

In the meantime, I found it pretty stunning that 41% of manufacturers are already deploying some form of software-as-a-service pricing based on having connected their products. This might take the form of charging for an hour of each engine’s run time or letting customers subscribe to a series of sensors for monitoring as a service. I know we’ve been talking about these possibilities for the last eight to 10 years, but this still feels like rapid adoption for such a slow-moving industry.

Beyond potential products, manufacturers are also looking to establish more of their computing at the edge, whether for data privacy and regionalization reasons, or concerns about the latency of data traveling from the factory floor to the cloud and then back again. Just 38% of respondents have already established their edge strategies for their own manufacturing operations, and only 29% have created an edge strategy for providing their customers with a local version of a connected service.

This makes sense. It’s hard to figure out what needs to happen at the edge. But the Microsoft survey does a good job of explaining some of the challenges manufacturers face when it comes to integrating their IT and operation technology (OT) data. Indeed, that combination of IT and OT is necessary when trying to determine what stays local and what goes to the cloud.

And while the IT-OT divide is closing, it’s still a thing. According to the survey, the biggest challenges to scaling out smart factories have to deal with OT security and interfacing with OT networks.

Finally, manufacturers are still figuring out what to do with digital twins. The concept of having a digital model of their equipment or entire factory operations is compelling, and roughly a quarter are already deploying some type of digital twin, but the rest are still trying to develop or don’t have a digital twin strategy. The most common reasons cited for the lack of a digital twin strategy were that it’s time-consuming — and expensive — to create digital twins. I blame that on a lack of standards.

What was surprising was how little manufacturers cared about augmented reality (AR); just 23% said they had an AR deployment in the works and 38% weren’t interested in one. Since Microsoft has invested heavily in the HoloLens, its pair of AR glasses for manufacturing, I’m sure these results aren’t welcomed. But I think it may simply be too early to ask the question.

Indeed, before we can get to AR we need to build high-quality digital twins and models for AR. Plus, we’re only now seeing partnerships, such as the one announced in June between Siemens and Nvidia, that will bring the computational oomph and manufacturing knowledge together to create a “factory metaverse.”  So we’ll have to wait.