IoT at the Edge
Showing results for 
Search instead for 
Did you mean: 

Exploring the Four Stages of an Industrial IoT Solution


I see every day, we’re surrounded by more and more intelligent devices and “things” – from cars and home monitoring systems to assembly lines, computers, and smart phones. These are all evolving into the Internet of Things (IoT), which has recently achieved celebrity status on par with earlier dot-coms, Big Data, and the cloud. Comprising troves of valuable datasets, the IoT is challenging organizations across all industries to tackle compute-intensive applications and tremendous workloads in order to remain competitive.

According to a recent report by MarketsandMarkets, the demand for hardware, software, and platform solutions and services specifically designed to support IoT solutions is expected to grow from $130 billion in 2015 to $883 billion in 2020, at a compound annual growth rate (CAGR) of 32.4% between 2016 and 2022. 

IoT Technology Market Size by Region 2022.png

  Source: MarketsandMarkets, Internet of Things Technology Market, 2016


Leveraging the proliferation of intelligent devices is giving businesses a choice: innovate, or be left behind. Increasing connectivity and adoption of HPC at the edge are helping enterprises shift compute functions out of the data center, right at the point of data capture. This streamlines access to relevant data, promotes lightning-fast analytics, and rapidly accelerates time-to-insight.

Several years ago, while working on an industrial IoT (IIoT) solution for a large energy company, it occurred to me that end-to-end solutions can be portioned into four distinct stages. This template has served us well as we develop IIoT solutions:

The Things - Primary Analong Data Sources.png

Source: Industrial Internet Consortium, 2015

Stage 1: Sensors and Actuation

Since analog data is the primary data type of the IIoT things, it must be captured via sensors. For example, a turbine in a power plant that generates electricity would have vibration sensors, as this analog data gives keen insight into the health and condition of the turbine. In reverse, control data and signals are connected to the things in order to control and actuate them. Robotic arms on a manufacturing floor, for example, require recalibration and refined control via actuators. This reciprocal relationship allows condition data to be sensed and analyzed so that real-time responsive control can take place.

Stage 2: Data Acquisition and Control Systems

This stage consists of systems that digitize the analog data and convert it into bits and bytes. Many times, controls systems, which direct Stage 1 actuators, are integrated with the data acquisition systems. The data flow from left to right then passes the digitized data to Stage 3. The control flow from right to left passes control information from Stage 3 to Stage 2.

Stage 3: Edge IT

The IT systems (compute, storage, networking, and associated software) which reside in this stage are not in the data center or cloud, but rather out at the edge. (See previous blog, The Big “Shift Left”: An Introduction to Edge Computing). While the cloud is the first off-premise architecture, the edge is the “other off-prem”. Analytics and prognostics on the things data can take place at the edge, closer to the point of data capture. This affords faster insights, real–time control, and reduced bandwidth strain and costs associated with transmitted all data back to a cloud. In addition, computing at the edge reduces security and data sovereignty concerns. I like to sum it up by saying, “Compute at the edge, accelerate insight”.

Stage 4: Data Center and Cloud

Most IIoT applications will employ the use of a data center or cloud, away from the edge and the things. I group data center and cloud together for simplicity, though in my experience of deploying clouds, a cloud is simply a data center that no one should be able to locate. That said, the disparate data center can be used for either all or part of the data analytics and control processing, hosting development platforms and global aggregation points for the edge through seamless communication and interconnectivity.

It is instructive to note that many IIoT installations are effectively “sensor to cloud” – from Stages 1 and 2 to Stage 4, skipping much of Stage 3. However, there is a growing demand to fill out Stages 2 and 3 as high-performance computing shifts out of the data center and to the edge. In a subsequent blog, I will illustrate other significant benefits of this four-stage architecture template, which involves mapping software, connectivity, and even vendor products for each stage.

HPE’s goal is to deliver accurate and immediate insights, affording customers edge and IoT solutions that are reliable, scalable, convenient, and efficient. Backed by cutting-edge HPC solutions, the seamless integration of these four stages are key to harnessing the power of the IoT.

I will be discussing these four stages and some real world examples during my speaking spot at the upcoming HPE Discover in London. Please see us there or contact us if you’d like to learn more. I also invite you to follow me on Twitter at @TomBradicich

0 Kudos
About the Author


Dr. Tom Bradicich is Global Head of Edge and IoT CoE & Labs,s at Hewlett Packard Enterprise. He and his HPE Labs team develop and commercialize advanced connectivity, compute, and controls software and technologies. Tom directs the HPE Edge and IoT Center of Excellence, which lead company-wide strategies, venture and M&A business and technical assessments, and the Channel-to-Edge Institute channel partner program.

Read for dates
HPE at 2019 Technology Events
Learn about the technology events where Hewlett Packard Enterprise will have a presence in 2019.
Read more
Read for dates
HPE Webinars - 2019
Find out about this year's live broadcasts and on-demand webinars.
Read more
View all