The industrial Internet of things (IoT) requires integrating information technology (IT) with operations technology (OT), and that can best be done at the edge, where the machinery being measured lives, rather than in a central data center or even in the public cloud. The value of much industrial IoT data decreases too quickly to tolerate the latency of transmitting it, even presuming that data networks can provide the bandwidth for the huge volumes involved. That means that 95 percent to 99 percent of IoT data will spend its life in edge computing facilities on places like off-shore drill rigs, factories and ships at sea, write Wikibon Chief Technology Officer David Floyer and Wikibon Chief Research Officer Peter Burris.

This has major implications for the architecture of IT in the next decade, the analysts say. It guarantees that many companies will need hybrid cloud environments that support those edge installations. It also implies that these edge installations will be highly automated, ideally run from a central site with limited direct attention from IT staff.

While industrial IoT covers a wide range of use cases, they break down into two categories for the purpose of IT design: static equipment such as buildings and mobile equipment, basically vehicles of various kinds.

The full report examines several factors impacting the development of industrial I0T. This includes advances in sensor technology, networking and ccompute.

It also discusses the main sources of business value from industrial IoT. One of the largest – and a major reason that most IoT data will be processed as close to the point of generation as possible – is making the most efficient use of expensive machinery, including identifying developing mechanical issues early to minimize equipment downtime.

Wikibon Premium users can read the full analysis here. To learn about subscribing, look here.

Image courtesy Wikibon