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

Widening the analytics lens with edge computing


By Lin Nease
IoT Chief Technologist, Hewlett Packard Enterprise

To see something big, you need to stand far away. Identifying an elephant, for example, isn’t easy if your visual field is limited to one small area.

data elephant.pngThe Data Elephant

Companies across industries such as manufacturing, energy, transportation, and automotive are well aware of this conundrum. For decades they’ve operated and maintained very large, complex equipment using relatively small data sets. But it’s not because the data was in short supply. On the contrary, complex machines—such as turbines, jet engines, oil rigs, and assembly lines—have long created troves of raw operational data.

Instead, the critical roadblock has been cost: The huge expense to collect, move, and analyze data forced most companies to prioritize the most critical information and use it for only the most critical insights. Thus, a holistic “full elephant” viewpoint into the operational environment remained out of reach.

The inherent possibilities of edge computing are changing the economy—and the possibilities—of Big Data utilization in manufacturing. But as companies forge ahead and create Big Data analytics engines at the edge, new questions are arising. Namely, can we ever have centralized intelligence from the edge?

The changing economy of Big Data analytics

Much of the hype surrounding IoT has focused on new types of devices and data—such as putting sensors in trash cans to give them a rudimentary intelligence, for example (“I’m full. Schedule a pickup.”). While this is exciting, from a data perspective it can be accomplished with just a few bytes and a miniscule amount of bandwidth.

Data use cases in manufacturing couldn’t be more different. They involve a tremendous volume of data requiring near real-time analysis. This is because the data is used for:

  • Performance—Preventing downtime that could lead to loss of revenue; avoiding damage to expensive equipment
  • Maintenance—Intervening proactively to maximize productivity across the lifecycle of machines
  • Safety—Preventing injury from malfunction or human error

In the case of a turbine, the volume of data produced can be gigabytes per hour—more than enough to make a wide area transmission unfeasible, in terms of cost, reliability, and increased risk of a data breach. Local processing, on the other hand, lowers the bandwidth cost by orders of magnitude, creates reliability that would be impossible in a multihop wide area network, and secures the data perimeter from external threats.

By changing the economy and boosting the speed of Big Data analytics, edge technology is helping manufacturing companies step back to see a wider view of their “data elephant”—and realize transformational improvements in cost and efficiency as a result. It is little wonder manufacturers have been early adopters of edge computing systems.

From cloud to edge and beyond

For many industries, “edge computing” is a new term but not a new idea. Private clouds were the first attempt at local processing and helped assimilate advances in digitization, automation, and control. But private clouds require data to be transmitted from the things; in addition, these virtual processing units are isolated and limited in capability, and therefore, data is underutilized.

Newer IoT technology, such as HPE’s Edgeline IoT Systems offerings, offers much greater scale and economy. Rather than maintain multiple expensive private clouds for each equipment cluster, Edgeline converged IoT systems allow companies to collect and analyze more data in an industrial environment.

Suddenly, organizations can afford to measure and monitor more than just critical data, leading to better control over operations because there is more input, better understanding, and greater speed. For example, a factory can develop machine-learning algorithms to predict consumption of raw materials and optimize its supply chain—in real time.

In short, edge computing is removing the islands of isolation in which manufacturing data has long been held captive.

From the whole elephant to the herd

Now that the analytics of industrial things’ data is more accessible, organizations can see through a wider lens. The full data elephant is coming into focus. But for many, this won’t be the last stage of the Big Data journey.

Computing at the edge gives individual site operators the ability to analyze lots more data locally. This enables you to focus your remote data center compute capabilities on greater Big Data analytics intelligence across multiple sites, or even further, your entire organization, in order to step back and recognize even more patterns and opportunities for improvement.


To learn more about the business drivers of edge computing, don’t miss our white paper HPE and IoT Compute at the Edge.

Related links:


Lin profile pic.jpg

Lin Nease is Chief Technologist, IoT at Hewlett Packard Enterprise. Connect with him on LinkedIn.

Empowering the Digital Enterprise to be more efficient and innovative through data-driven insights from the Internet of Things (IoT)
0 Kudos
About the Author


Online Expert Days - 2020
Visit this forum and get the schedules for online Expert Days where you can talk to HPE product experts, R&D and support team members and get answers...
Read more
HPE at 2020 Technology Events
Learn about the technology events where Hewlett Packard Enterprise will have a presence in 2020
Read more
View all