Harnessing Data for Digital Transformation
Organizations considering IoT projects should understand how to collect, store and analyze data to gain insight and automate processes.
by | Kym Gilhooly
Kym is a freelance journalist who specializes in business technology and is a frequent contributor to the CDW family of technology magazines.
For organizations in every industry, data is coming in massive waves. Busy IT teams and line-of-business leaders have two options: either use this data to improve operations or drown in the tsunami and fall behind the competition.
The sources of this tidal wave are varied. More devices are generating more data than ever before, and it needs to be captured, stored and analyzed by organizations that want to make use of it. A 2016 Business Insider study estimates that 34 billion devices will be connected to the internet by 2020 — all generating data. These include 10 billion computing devices, such as notebooks, tablets and smartphones, as well as 24 billion devices connected via the Internet of Things (IoT).
To make use of this data, organizations must work through a digital transformation. Those that do so are setting themselves up to reduce operating costs and increase revenue by expanding their offerings or market presence.
When IT and OT Systems Meet
Today, networking technologies and sensors enable organizations to collect data from virtually everything: medical and mobile devices, appliances, cameras, wearables and vehicles, which can be assembled in an array of combinations to create smart homes, buildings and even cities.
But leveraging data from IP-enabled IoT devices is just part of digital transformation, says Link Simpson, IoT practice lead for CDW. At its core, he says, IoT is the convergence of IT and operational technology (OT). Data from these systems can be combined with and analyzed by IT systems to yield benefits such as remote monitoring, automation, streamlined processes and operational insights.
Among OT systems, organizations are leveraging data created by new sensors coming online. But smart organizations also use data from legacy systems such as supervisory control and data acquisition networks, programmable logic controllers and distributed control systems (in manufacturing environments, for example). In addition, they employ building management system components such as heating, ventilation, air-conditioning, power, lighting, elevators and security systems.
“This isn’t a rip-and-replace,” says Simpson. “We need to tie in all these existing systems that required significant investment and still have a long life.”
Many of these are stand-alone systems, says Simpson, so they first need to be connected to an IP network so they can sense each other, and then converted to a common protocol so they can communicate. IT and OT teams work together with service providers to connect stand-alone systems and merge operational networks with the enterprise network. This not only connects these systems through a single network, but it also enables stakeholders to leverage data from enterprise resource planning, customer relationship and other critical applications.
Pushing Data to the Edge
Though numerous cloud providers offer analytics and data storage services, Internet of Things (IoT) deployments often have processing requirements that don’t mesh well with the cloud. Edge computing provides a useful solution to this problem.
Some data sets need to be processed in near-real time, and latency issues can slow the retrieval of data stored in the cloud. The edge computing model conducts data processing functions at the edge of a network rather than sending data for processing in a cloud or a data center. “Sometimes half a second is too long, so you’ve got to get your compute very close to the edge,” says Link Simpson, IoT practice lead for CDW.
With the mountains of data that IoT devices generate, he advises putting some computing, decision-making and triggering functions at the edge. For instance, video cameras are among the most powerful sensors in operation today whose output enhances analytics. But if a camera takes video continuously for surveillance purposes, streaming all the data it creates and storing it all in the cloud creates an onerous burden that organizations should avoid. “No — you wait until there’s an event, and you send that up,” says Simpson.
Edge computing is the best approach from both a bandwidth and storage standpoint, says Simpson, “You balance edge computing with what can be done in a centralized location,” he adds.
IoT? It’s Complicated
Connecting multiple systems in an IoT deployment is a complicated undertaking, requiring systems integration, network convergence, a robust wireless deployment and increased storage capacity, as well as expanded cybersecurity efforts. To manage these complexities and the cultural differences between IT and OT teams, many organizations turn to professional service providers with significant IoT expertise.
Despite the attention IoT has received, most organizations are still planning or in the early stages of implementation for their projects. Experts suggest that organizations start with projects that have simple use cases, primarily internal, that they can then build on.
In a 2016 Gartner survey of U.S. organizations planning or engaged in internal IoT projects, 81 percent said their focus through 2018 would be on optimizing operations in areas such as workflow, supply chain and inventory management. For organizations planning or starting external IoT initiatives, 72 percent expected to concentrate efforts through 2018 on smart connected products that allow their enterprise systems to receive data from customers.
Streamline, Automate, Optimize
Despite its complexities, IoT holds great promise for organizations that develop a smart strategy. With networked sensors configured for specific use cases and generating data for analysis, they can identify areas ripe for operational efficiencies, new sales channels and premium product tiers.
Enabling remote monitoring: With the ability to put smart sensors on or in just about anything (including people), IT and OT staff can remotely monitor resources and take action based on the data. For example, healthcare providers can remotely monitor devices such as insulin pumps or subcutaneous vital sign sensors; companies with field service teams can monitor the safety of workers involved in high-risk activities; and oil and propane providers can check tank status remotely, drastically reducing costs by avoiding unnecessary travel.
Gaining insight: By capturing new data generated by IoT systems, organizations can apply analytics and turn this information into actionable insight. It’s the combination of analytics with IoT innovation that delivers on digital transformation investments, says Simpson.
Not everyone wants to build their own analytics solutions. CDW advises customers to deploy best-of-breed, industry-specific analytics tools, so they don’t have to start from scratch. “With solutions designed for retail or manufacturing, they get 80 percent of the functionality they need off the shelf, and we can configure them to deliver the rest,” he says.
Streamlining processes: Every organization has its own business processes and workflows and is continuously working to optimize them, says Brad Shimmin, service director at Current Analysis. To accomplish this, they instrument these processes. A sensor that records a person entering a store, for example, generates data that a retailer can use to improve its efforts at sales, advertising and customer service.
To that end, IoT data aids optimization by filling in the gaps in existing processes, where some aspect couldn’t be instrumented. “IoT isn’t new — it’s what we’ve been doing, but a lot better, because we have a more complete picture,” Shimmin says. “You can see data in real time, and make decisions closer to the point of inflection.”
Automating operations: Automation is a popular IoT use case throughout a variety of industries, from smart building systems to retail stores to shop floors.
Smart elevators, in some cases integrated with calendaring systems, automatically adapt to a building’s events and employee schedules for more efficient transportation and energy savings. On shop floors, smart sensors predict equipment failure and send alerts that trigger automated break/fix responses. Automation in supply chains can optimize shipment logistics and inventory management.
Identifying automation opportunities and taking action, however, requires that organizations be skilled at building predictive models and analyzing complex datasets, or that they hire outside contractors for guidance.
IoT: A New Threat Vector?
Some Internet of Things experts claim that “IoT” instead stands for the “Internet of Threats.” Indeed, IoT brings huge opportunities for innovation and automation, but it also comes with significant risks.
For example, IoT manufacturers may sell products that do not include hardwired authentication credentials, or they may fail to advise buyers to change default password settings. This omission led to a 2016 situation in which the powerful Mirai malware pulled together hundreds of video cameras and routers to create the monster botnet that carried out a massive distributed denial of service attack.
“Security has to be considered at every step when developing an IoT strategy,” says CDW’s Link Simpson. “You can’t secure a lot of legacy operational technology systems through standard cybersecurity means.”
Organizations should update their security policies to defend connected IoT devices. One important tactic is to employ network segmentation — through solutions such as firewalls, zoning, intrusion protection systems and virtual local area networks — in order to prevent IoT devices from accessing the enterprise network. IT teams also should review identity and access management solutions as well as user access privileges.
Manish Jiandani, Splunk’s director of solutions marketing, advises IT teams to take a “security first” approach when mapping IoT connectivity and integration phases.
Dashboards Deliver on Data
To gain insights from the data they gather, organizations must carefully employ analytics tools. Some platforms — such as Splunk — integrate analytics software with data preparation, governance and security tools. Splunk focuses on operational analytics, so its tools complement automation and orchestration efforts. The company’s platform can collect machine data in real time from a multitude of sources, including applications, IoT sensors, web servers, databases, networks, virtual machines, mobile devices and mainframes.
“Customers can create the traditional dashboard to monitor the health of the infrastructure, performance of applications, and do basic reporting and analysis,” says Manish Jiandani, Splunk’s director of solutions marketing, business analytics and IoT. “On the other end, we have a machine-learning toolkit with built-in algorithms, which allows customers to take IT and business data and build a model customized to their specific requirements.”
For organizations getting swamped by digital transformation, dashboard-driven analytics solutions enable visualization that helps decision-makers understand the massive amounts of data they generate.
Call us at 800.800.4239 to set up a consultation with a IoT expert.