As senior manager of asset-performance management (APM) product marketing with

AspenTech, Robert Golightly knows the importance of machine maintenance and the value in predicting when things will go south. Here he shares his perspective on getting real, reportable results that inform real, preventative actions.

Smart Industry: What are the biggest challenges for companies in demonstrating ROI results?

Robert: The biggest challenges are threefold.  

The first challenge is to clearly identify the objective of the IoT project. Merely having access to IoT technology doesn’t equate to value. Organizations that opt to digitize everything in their path will get a boatload of data, but no clear understanding of how to use that data. That’s why it’s important for organizations to identify, upfront, where their biggest challenges are and digitize only for that—solving for a specific problem. Data is not the answer; it’s what organizations actually do with the data that matters. That’s where they’ll find their ROI.

Secondly, prioritizing areas of impact is a must to demonstrate clear and compelling ROI. The only reason any industrial company embarks on an IoT project is to drive higher levels of operational excellence. Those priorities can be ensuring safer environments, more efficiency, more profitable operations, improved sustainability (energy management, for example) to bridging the talent gap and driving improved organizational excellence. For example, we’ve seen results from real-world applications of prescriptive-maintenance software in the oil-and-gas industry. At a large North American refinery, our prescriptive-maintenance software gave eight weeks advance notice on a compressor failure—a full seven weeks before its state-of-the-art vibration system detected the impending breakdown. Prioritizing the area of impact—here, key, costly plant equipment—makes for a clear ROI analysis. The costs of emergency shutdown, loss of available equipment to produce product and unhappy customers (plus the potentially disastrous EH&S implications) will add up significantly compared to the investment in a quickly deployed, low-touch analytics and machine learning prescriptive-maintenance application.

Thirdly, quantifying improvement targets will make or break any ROI exercise. IoT ROI comes from clarifying what you are applying the technology for and then comparing the returns from new, emerging tech to older…those tried-and-true methods. Any solution needs to fit into industrial companies’ workflows and legacy-technology infrastructure. That’s where the ROI becomes most clear, when new tech drives high and sustainable levels of operational excellence.

Smart Industry: In digital transformation, is it possible to avoid “rip and replace” with old equipment or legacy systems?

Robert: Improving operational performance to increase return on assets and manage risk is nearly impossible without connected assets. Here are the challenges: companies encounter connectivity issues in the form of out-of-the-way locations that lack even basic internet services, old equipment not yet sensor-ized, a mash-up of incompatible protocols, OEMS and Programmable Logic Controllers (PLCs). In the industrial world, these are “stranded assets;” up to 40% of industrial assets fall into this category.

Connecting assets, aggregating data and performing advanced analytics delivers the data-driven insights needed to improve performance (as measured by more reliable assets that produce more, break down less and cost less to fix). The IoT has the potential to offer a huge competitive advantage for companies, who can then use those real-time operational insights to make faster and smarter business decisions and reduce operating costs. This is all, in large part, made possible by one phrase that bears repeating: the edge.

The ROI on the IoT comes from driving reliability enterprise-wide without ripping and replacing existing infrastructure—with drop-in edge-computing solutions that can reduce deployment costs by up to 70% and can be implemented in days, not weeks. Connectivity at the edge simplifies IoT applications and seamlessly integrates and interoperates with legacy systems. Analytics at the edge eases the processing strain on the entire network; buffering, data aggregation, data compression and analytics on the edge optimizes and overcomes the connectivity challenges present in oil fields, refineries, mining sites, power facilities and other industrial environments.

To get the best ROI, consider the costs of cellular or satellite networks, as well as the storage costs of the cloud; then investigate edge connectivity software that supports Windows and LINUX edge devices using standard industrial protocols, such as MQTT, OPC-UA, OPC-DA and Modbus. Working with what you have to make digital transformation work for the business is where ROI and IoT intersect.

Smart Industry: How should organizations approach demonstrating prescriptive-maintenance savings to management/executives?

Robert: The World Economic Forum estimates IoT could contribute $20+T to global GDP by 2020. You’ll need more to convince the higher ups that prescriptive maintenance is adding real, measurable value, though.

To project and report savings from prescriptive maintenance, convincingly, to your management, think about the shared operational-excellence goals everyone is trying to achieve. Most IoT applications involve customers trying to solve complex problems that involve many data variables with intricate relationships that demand the use of big-data tools and large amounts of compute power. There are many applications for AI, multivariate analysis, pattern recognition, event detection and machine learning. Get right to the bottom line instead. How far in advance was the failure detected? These examples below project and report one (or two) compelling numbers that show value.

• In a drilling operation, prescriptive-maintenance software correctly detected calibration errors on drilling joystick operations that had gone unnoticed. Prescriptive-maintenance software provided two to four weeks’ warning of impending failures on top-drive, mud pump and draw-works components.
• A multinational mining company used prescriptive-maintenance software to significantly improve production uptime with an average time-to-failure of 40 days on a pump. An oil-and-gas company had experienced unexplained breakdowns on numerous compressors. Prescriptive-maintenance software provided notice seven weeks earlier than the state-of-the-art vibration-analysis system.

We heard from one European customer that for most companies, 15% gross margin losses are attributable to unplanned versus prescriptive maintenance. Even best-in-class approaches 4–5% losses.

Smart Industry: What is the value-add of prescriptive analytics for organizations?

Robert: Prescriptive analytics offer the best way to use reliability for competitive advantage. Nothing hurts asset-intensive businesses more than unplanned downtime. Prescriptive maintenance is one of the best ways companies can apply advanced analytics and machine learning to a clear and present problem. The National Association of Manufacturers in 2016 suggested worldwide, manufacturing is a $14T year business with a big problem—10% of manufacturing losses are from equipment breakdowns costing $1.4T each year. We call our own prescriptive-maintenance software “the Science of Maintenance.” This concept centers around turning traditional maintenance into a data-driven initiative that improves reliability by applying IoT tech to improve operational excellence.

Smart Industry: Provide examples of organizations that excelled in showing IIoT ROI?

Robert: The Saras oil refinery used AspenTech prescriptive-maintenance software to detect impending failures with 91% accuracy, with 30 days of advance warning. Lower maintenance costs were a benefit—but the biggest value was reliability as a clear driver of operational excellence in the areas of safety, efficiency and profitability.

Specialty-chemicals company Borealisgained longer lead time detection of repeating failures, including 27 days advance notice for central valve failure, transfer learning (‘inoculating’ similar assets with software-defined failure signatures to scale the benefits of prescriptive maintenance enterprise-wide) and the ability to capture fast-moving failures before they caused major damage.

Smart Industry: What are some best practices to improving and demonstrating IoT ROI results?

Robert: Apply knowledge, not just technology. Capitalizing on the volumes of sensors deployed and data collected is more about the application of technology than the technology itself. Knowledge in the form of data science embedded in software transforms data into tangible and repeatable benefits for customers. The result is sustainable software applications that leverage machine learning, knowledge automation and systems-level thinking to achieve digital transformation and drive operational excellence.

Reliability is an industrial-business-model disruptor that addresses the bottom-line-oriented operational functions of asset and capital-intensive industries. There’s a huge opportunity to deliver greater reliability and greater ROI on IoT via prescriptive-maintenance software to many industries including metals & mining, pulp & paper, power—like the energy and chemicals companies, these verticals all have critical, costly-to-fix equipment that shut down the business when they break. With low-touch prescriptive-maintenance software, getting ROI from the IoT is now within reach.

Want more on smart maintenance? Find our latest Technology Report—"Turning Downtime Upside-Down”—by clicking here.