The last thing a manufacturing IT leader needs is another issue to keep them awake at night. But for many, that’s exactly what the Internet of Things (IoT) is turning out to be.
Let’s look more closely at why IoT has become a source of anxiety within so many manufacturing firms—and how engineered systems infrastructure can shift the focus back on IoT as a massive source of value for manufacturers of every size, and an essential first step towards smart manufacturing.
The Deeper Concerns Behind IoT Worries
Like so many manufacturing technology challenges, the issues with IoT seem pretty clear-cut at first, but a closer look reveals a more nuanced story.
Three findings from a recent survey of IoT stakeholders explain, in a nutshell, why manufacturers experience IoT-related angst:
Why does IoT data cause so many headaches? There are two main issues in play:
Manufacturing systems and other devices have long been designed to generate data used for management, monitoring, and maintenance tasks. These capabilities are the foundation for what is called operations technology (OT): a class of systems designed to monitor physical devices, processes, and events within a manufacturing or other business environment.
Many of the OT systems in use today, however, have pre-internet origins—a time when storage, computing, and other resources were relatively scarce and very expensive; and when closed, proprietary standards and protocols were the norm.
These largely on-premises OT systems generate data that tends to live in siloes, so it’s isolated and difficult to access, and even more difficult to integrate effectively.
Turning IoT Challenges into Opportunity
It’s not all bad news: Manufacturers do have easy and economical access to technology that cuts IoT challenges down in size.
The key is the emergence of engineered systems: a stack of hardware and software infrastructure that is architected, integrated, tested and optimized to power business applications, technologies, and decisions. For a manufacturer, an engineered system combines the cost and performance advantages of modern, commodity hardware with cutting-edge software capabilities, advances in integration and process automation, and a simple, single-vendor approach to make it all work at peak performance.
Let’s consider some key points in a typical IoT scenario where Oracle Engineered Systems in particular can make a difference for a manufacturer:
Oracle Engineered Systems products such as Big Data Appliance along with key big data analytics offerings allow manufacturers to achieve this level of performance and analytical insight. These combined solutions help manufacturers aggregate massive quantities of data from IoT environments that may encompass thousands of devices with multiple data sources, and billions of discrete data points at any given time. Big Data Appliance delivers the infrastructure layer to help acquire, aggregate, store, and process data of virtually any volume and any type in an open Hadoop environment. These capabilities, in turn, give the big data analytics offerings a solid, scalable, and reliable foundation for delivering business-critical analytics solutions.
IoT Impact: Bigger Benefits for More Manufacturing Applications
When used effectively, IoT data and analytical insights supported by powerful, cloud-ready infrastructure can enable classes of applications to be far more valuable:
OT Isn’t Going Anywhere
OT is still valuable, and indeed an essential tool for maintaining efficient and productive on-premises manufacturing operations. The goal now is to maximize the value of OT, placing it within an integrated set of capabilities—all of which call upon IoT data, analytics applications, and streamlined infrastructure.
Manufacturers are now challenged with integrating these applications with OT capabilities and supporting it with modern infrastructure—all while operating in a high-pressure, high-performance environment. Oracle Engineered Systems overcomes this challenge with “co-engineered” infrastructure products that are optimized to work with even the application layer and improve performance. This approach supports scalability and can be tailored to work with a range of deployment models: from traditional, on-premises data centers to public-cloud configurations.
Turning IoT Data into Practical Analytical Insights
This level of “performance under pressure” is very common for organizations that want to get value from their IoT environments. A recent example, involving the semiconductor division of a multinational electronics manufacturer, illustrates this point.
The key challenge for this firm, which maintained R&D centers in the United States and Asia, was a statistical analysis process that simply wasn’t up to the task of getting useful insights from a massive set of IoT resources: 500,000 sensors and 3.5 billion data points spread across facilities on two continents. The firm was determined to upgrade its IoT data collection and analysis capabilities; to gain the right insights to improve product quality and equipment performance; and ultimately to boost manufacturing yields—potentially a major advantage in a highly competitive industry.
An Oracle Engineered System for big data analytics turned out to be an ideal fit for this firm’s IoT needs. The engineered systems approach didn’t just hold up the massive data volumes involved; it actually enabled near-real time data analysis that could identify the root cause of equipment failures as they happened. The Oracle solution also integrated the firm’s IoT data streams with analytics tools that revealed previously hidden patterns and trends and helped to predict the results of manufacturing process changes.
By capturing and unlocking the insights within its IoT data, the firm achieved its goal of higher manufacturing yields. In the process, it identified some important new methods to enhance product quality, even as it used efficiency gains to reduce operating costs, and achieved incremental sales and revenue gains.
Airbus Stays Ahead of Fight-test Data Challenges
In many cases, as the previous example suggests, it’s not enough simply to grind through these types of large-scale analytical tasks. Manufacturers are engaged in a perpetual race against the clock; they need solutions that keep them ahead of competitors and that avoid creating chokepoints in existing manufacturing processes.
Another case history, this time involving aircraft manufacturer Airbus, drives home the value of using the Engineered Systems approach to solve analytical problems where time is money - and where even minor performance hiccups can impose unacceptable delays.
As Airbus ramps up production—it expects to produce 30,000 new planes over the next two decades—it must also scale and streamline its flight-test processes. Today, a typical test flight lands with about 2TB of data, providing a source of potentially critical insights into aircraft performance, efficiency and flight safety.
Airbus uses the Oracle NoSQL Database running on the Big Data Appliance to ingest this test data, store and manage it, and make it accessible on-demand to Airbus analyst teams. The Big Data Appliance gives Airbus a robust infrastructure that moves test data exactly where and when it needs to go—allowing the company to shave 30% off its average testing time even as it continues to scale its manufacturing and flight-testing processes.
Could Airbus build its own big data infrastructure solutions? Of course it could. But the Airbus management team knows its capabilities are best applied where they are most valuable: testing and improving their aircraft systems, not building big data infrastructure.
A Better Way to Benefit from IoT Insights
We all know just how fast technology innovation moves today. Most of us are also familiar—typically from first-hand experience—with the pain that often results when amazing new capabilities run up against legacy systems and data. It’s a dilemma that is on full display when manufacturers see the potential within their IoT data but experience the realities of dealing with legacy OT environments.
It doesn’t have to be this way. Engineered systems, deployed in ways like the ones I discussed here, give manufacturers a simple and affordable way to cut through the confusion and complexity, and to turn IoT data into revenue-impacting insights.