Skip to main content
Exploring what’s next in tech – Insights, information, and ideas for today’s IT and business leaders

Why your business needs data insights from the intelligent edge

The benefits of analyzing data closer to where it is generated, at the edge of the corporate network, include faster insights, and more secure and compliant data.

The Internet of Things (IoT) is nothing short of a revolution. Suddenly, vast numbers of intelligent sensors and devices are generating vast amounts of data that contain potentially game-changing information. The challenge for IT leaders is to gather insights from that data rapidly, without breaking the bank, opening up security holes, or running afoul of compliance mandates. It's a tall order.

With earlier data business intelligence projects, organizations shipped data to a central data warehouse for processing in order to glean strategic insights. Big data projects were similar, tossing large amounts of data of varying types into a data lake to be pored over later.

But now, with IoT data being churned out in bulk by sensors in factories, warehouses, and other facilities, many organizations are finding it makes much more sense to analyze it close to where it is generated: at the edge of the corporate network. Benefits include faster insights and data that is secure and compliant. Best of all, you avoid the high costs of sending gigantic volumes of data across large distances.  

"It's like the human nervous system. If you touch something hot, your hand responds before sending the information to your brain," says Sam Ransbotham, associate professor of information systems at the Carroll School of Management at Boston College.

When data networks handle sensor information in the same way, similar benefits result. "The old way was to put everything in one place and use a big network to move the data. But that doesn't make a lot of sense anymore," says JR Fuller, worldwide business development manager for IoT Edgeline systems at Hewlett Packard Enterprise.

There is no mistaking the trend. In a recent IDC survey, 97 percent of businesses surveyed said they collect data at the edge. Of those businesses, 54 percent transmit that data to a central enterprise location, while 43 percent process it at the edge.

IoT insights at the edge

In a factory, sensors on a machine that makes pumps might generate hundreds of bits of data each second. Over time, patterns in that data—showing vibrations, for example—can tell the story of a manufacturing defect about to happen. Sending the data across a wide area to be studied is costly and time consuming and might take so long that preventive action can't be taken. In contrast, when plant managers understand what's going on while the machine is working, they can intervene before the defect occurs. Thus, processing the data locally lowers costs while increasing productivity.

Keeping things local yields security and compliance benefits as well. Every IoT sensor represents a potential point of compromise. But should a piece of malware infect a system, keeping the data from traveling across a network prevents the malware from wreaking havoc with data at corporate headquarters. And when the data at the edge is personally identifiable information (PII), a typical case in the financial services sector, preventing it from leaving a country can keep a company from running afoul of data privacy laws, particularly in Europe.

But it's not an either-or proposition. Data at the edge can be acted on instantly as needed, and a subset of it can be sent to a central site for aggregation with data from other sources, to yield insights that can guide long-term strategy. This practice is reflected in IDC's numbers. Of the 43 percent of businesses that process data at the edge, 29 percent collect and process data at the point of creation. Another 14 percent collect and process some data at the point of creation, while transmitting the rest to a central site.

"The sensors have to be aggregated into a system of intelligence. You're going to have to do some of this at the edge, decide how much of this data is really worth something, then dictate how fast that data is going to go to a central location," explains Vernon Turner, senior vice president and research fellow for the Internet of Things at IDC.   

The IoT 2020 IT Infrastructure Readiness Indicator: IDC believes that despite massive investment in IoT, much of the impact of this transformation is yet to be felt. Learn how to approach IT infrastructure readiness.

Edgy business models

The case for IoT edge analytics may seem like common sense. But there's more at stake, asserts one analyst. "In the IoT world, it's really important that companies think about new ways to use this connected technology. It's not about doing the things we have done in the past better. The goal should be to find new ways of doing things," says Zeus Kerravala, founder and principal analyst at ZK Research.

Indeed, once started down a path along the edge, organizations are discovering new business models that are anything but common. In many cases, the new model represents a change from selling products to selling business results. From the customer's point of view, it's a shift from purchasing an item as a large, one-time capital expense, to the smaller ongoing payments of an operating expense. It's a difference that many organizations find attractive.

Pitney Bowes is moving beyond its traditional business of selling mail processing equipment and services to selling productivity itself. In partnership with General Electric, the company has launched a new offering called Clarity Solutions Suite, which closely tracks equipment performance both at the edge and at a central site, so customers pay for the performance of the equipment rather than the equipment itself.

Pitney Bowes' services encompass multichannel contact with customers—including text, email, traditional mail, and telemarketing—through partners. Traditional mail services include assembling, weighing, addressing, and shipping envelopes and packages. Sensors on the machines that perform these tasks generate information that is handled both at the edge for immediate response and at a central site for further analysis.

"We have been creating millions of data records every minute and every hour for a while, but people haven't been leveraging them. Now we're harnessing the data that's running through the machines," says Grant Miller, chief operating officer for document messaging technologies at Pitney Bowes. In the mail industry, where the data is processed and stored is important, he points out. Names, addresses, credit card numbers, and medical information must by law remain in a single physical location, and that may mean never leaving the edge.

With Clarity, both the customer and Pitney Bowes share a common view of key performance indicators and service-level agreements so they can track progress together. When a gain in productivity is realized, the customer shares the profits with Pitney Bowes. "If the target is to improve production by 3 percent, they send us a share of that amount when the goal is reached," Miller explains. Clarity Solutions is being offered across a variety of industries, including banks, telecommunications providers, insurance companies, and government agencies.

Experimentation adds value

Precision farming is another field that's ripe for IoT at the edge. Sometimes called precision agriculture or satellite farming, the practice is based on GPS information in combination with data from sensors on the ground that measure moisture, nitrogen levels, terrain, and other factors. The data is analyzed to generate insights for crop optimization, then instructions are sent to seeding and irrigation machines to fine-tune cultivation tasks in the field.

Inmarsat, a London-based satellite network company, has built a platform on its broadband global-area network that enables farmers to make decisions at the edge in real time based on a local repository of data, while higher-level decisions are made at the farm's head office.

"The analytical processing is through a combination of processing at the edge and in the cloud. The control of pumps, irrigation systems, and fertilizer units are done in real-time at the edge, while data is fed to the cloud for business intelligence, data validation, and control validation," explains Paul Gudonis, president of Inmarsat Enterprise.

"Because edge analytics provide reliable and constant processes, efficient offline decisions are made in near real time. Video surveillance analytics at the edge provide significant savings and instant user gratification in terms of supporting business processes," Gudonis adds.

Precision farming is still a work in progress, as new insights and approaches are being found by recombining information in new ways. "Value also comes from education and experimentation once an IoT infrastructure has been created, by testing new use cases and testing business information in an iterative manner," Gudonis adds.

Intelligent edge gotchas

While the payback from edge analytics can be game-changing, there are plenty of obstacles that could derail even the best efforts. The very nature of IoT devices is cause for concern: There are lots of them; they tend to create new relationships, sometimes with third parties; they are often placed in dirty and dangerous locations; and they may create security vulnerabilities.

"Security in general is an issue, and everything about IoT makes it worse, not better," says Ransbotham. Compounding the problem, Ransbotham's recent research has found that only about one quarter of managers are concerned about IoT security. "It makes me wonder about the other three quarters," he says.

Another gotcha: managing expectations. It is easy to underestimate the effort and overestimate the immediate benefits of an IoT project. "People expect to see the ROI right away, but patience is required," says Laura DiDio, former research director for the Internet of Things at 451 Research.

Despite the allure of an ambitious IoT project, some implementations just aren't meant to be. In a fall 2016 article in the MIT Sloan Management Review, Ransbotham notes that the city of Amsterdam considered installing smart LED bulbs in all of its streetlights, expecting to save money by dimming the lights when pedestrians and vehicles were not near them. But city authorities scrapped the plan when they found out that workers would have had to swap out 150,000 bulbs, an impractical task in itself, even before factoring in the cost of the wirelessly enabled smart LED bulbs and the network devices needed to talk to them.

The problem faced by the Dutch city is typical of an IoT pitfall Ransbotham calls "diseconomies of scale." The number of sensors—and the fact that they must be managed, maintained and sometimes upgraded—must be factored into any project, he cautions.

Taking control of IoT data

To overcome the perils that lurk in the unexplored territory of IoT and edge analytics, Fuller recommends addressing four key challenges head on:

  • Manageability: It's easy to manage IT resources in a data center, but when sensors are all over the world, it's much harder to keep track of them. Standard protocols are emerging, however, and using the ones that are relevant to your implementation is key to getting a handle on a plethora of devices.
  • Security: Every sensor and edge device increases vulnerability to some degree. Keeping IoT data at the edge where it belongs is a first step. Beyond that, the way to go is a defense-in-depth approach, in which physical, technical, and administrative controls are placed at strategic points across the IT infrastructure.
  • Data governance: It is essential to know what data is collected, where it is stored, and where it is being sent. Depending on the data and the rules governing it, geofencing may be necessary. Governance policies should be applied to data stored at the edge. Useless data should be discarded quickly. Data and metadata for analytics might be held much longer but should be deleted when it's no longer useful.
  • Change management: People and organizations need to adapt to the new approach of working at the edge. An adaptive and agile organization has a much better shot at extracting value from IoT data—and moving on to establish a new business model should one emerge.

All in all, the upside of a successful IoT initiative spurred by edge-based analytics is too great to ignore. From smart cities to pay-as-you-go services, conventional business models are being disrupted. Change is spreading across nearly all industries, and organizations must move quickly or risk being left behind their faster moving peers. IT leaders should go into the new world of IoT with their eyes open to both the inherent challenges they face and the new horizons that are opening up.

Data at the intelligent edge: Lessons for leaders

  • Scout it out: Figure out whether you have IoT data that's worth gathering and analyzing, and whether deploying a host of sensors is likely to yield a return.
  • Keep calm: Bridle your expectations. And be on guard for diseconomies of scale.
  • Stay safe: IoT means a new security paradigm. Figure out where the vulnerabilities are and apply defense in depth.  
  • Consider governance: Regulatory compliance often requires having your data stay put at the edge. 

Related link:

The fast changing world of data & analytics

This article/content was written by the individual writer identified and does not necessarily reflect the view of Hewlett Packard Enterprise Company.