Servers: The Right Compute
Showing results for 
Search instead for 
Did you mean: 

Taming the IoT data tsunami with edge computing


Guest blog written by JR Fuller, WW Business Development Manager for IoT Edgeline Systems

JR Fuller headshot.jpg

The Big Data tsunami fueled by the Internet of Things (IoT) just keeps growing, fed by inputs ranging from jet engines to oil rigs to warehouse forklifts. Transmitting terabytes of data from their source to a central repository before analyzing it can be extremely time-consuming, and in many industrial settings such latency can have serious repercussions.

For example, if you’re the supervisor of a nuclear power plant, you don’t want to wait to find out that something’s going wrong. If temperatures start to spike or a piece of equipment begins to misfire, you want to know right away and take immediate action. Having to send that information back to a central reporting facility could introduce a disastrous delay, but by analyzing streaming sensor data close to the source—at the “edge” where the data is generated—you’re able to make quick decisions and act in near real-time.

The capability to analyze IoT data is becoming an increasingly important competitive differentiator. According to a recent ESG Research Report, 45% of enterprise IT professionals surveyed believe the IoT will improve operational efficiencies, and 26% feel it will help them develop new business models. But how do you manage IoT data in a way that lets you realize those goals?

Don’t clog your network with data sets—analyze data at the edge

JR Fuller blog pic.jpg

Not every situation calls for immediate data-processing capabilities. That decision depends on several factors, including time to insight, depth of analysis, and the amount and type of data. The faster you need the information (time to insight) to make decisions, the closer you want the analysis to the data source. The more analysis required to arrive at an outcome (depth of analysis) typically drives the data to a centralized compute center, whereas simple processing and control is preferred at the edge. If you’re in an environment such as a power plant or manufacturing facility that requires you to continuously monitor your systems, you’re likely generating a lot of data. This makes edge analysis essential.

Take video surveillance analytics as an example. These analytics often generate huge amounts of data, and you might not be able to transmit all the video generated to a central location. Transferring all of that video information to the cloud or a data center for storage and future analysis would consume a significant amount of bandwidth. Furthermore, it could potentially increase transmission costs, especially if the amount of data you move determines how much you pay. You may also wish to reserve that bandwidth for other uses, such as email or other applications.

By understanding your data and your organizational needs, you can deploy compute power at the edge when appropriate. This allows you to make smart decisions on site to compile and filter your data. That way, you can transmit just the metadata needed, rather than saturating your wide area network connections with an entire data set.

Weighing data size and security

A number of additional factors can help you determine where best to analyze IoT data. The scope of data defined for analysis is critical. You’ll want to weigh the need for real-time insights derived at the source of input via edge compute resources against a desire for longer-range and more detailed analysis, which can be handled in a data center or the cloud.

Data security and integrity is another significant consideration. The more data you transmit, the higher the odds are that your data will be vulnerable to an attack. You also raise the risk that data will be corrupted during transmission. Analyzing data at the edge and sending only a select portion back to your data center can help safeguard important information.

Depending on your industry, you may also need to consider various compliance issues. Sending data across borders can complicate operations for multinational organizations. In some situations, local geo-fencing laws may prohibit data from being transmitted outside designated boundaries.

Companies are eager to start reaping rewards from data insights gleaned from the Internet of Things. To take advantage of IoT opportunities, you need to map out a data strategy that makes the best use and combination of your infrastructure resources, whether they’re edge, cloud, or data center.

Read the ESG white paper “HPE and IoT Compute at the Edge” for more insights on how best to deploy IoT data processing capabilities.

Follow JR Fuller on Twitter at @JRFuller321.

Related links:

Visit our HPE IoT website

Follow us on Twitter at @HPE_IoT

Infographic: Harnessing the power of IoT

Video: HPE helps DS Virgin Racing team improve racing performance

Article: IoT turns ordinary data into extraordinary business outcomes

0 Kudos
About the Author


I am a Senior Manager managing external content and social media for HPE Servers Awareness. Stay tuned for topics on Mission Critical Solutions, Core Enterprise and SMB Solutions, Next Gen Workload Solutions, Big Data and HPC, Cloudline and HPS Options! Follow me @RubyD_Nich

Johannes Horneck

Great article. If you are interested: Meet HPE at the big HMI fair in Hannover, GY end of April.