How Wireless Tech is Changing Predictive Maintenance
January 28, 2020
Thanks to wireless connectivity, sensors, the IoT and AI, industrial plant operators can save time and reduce costs by remotely monitoring critical assets.
In factories around the world, millions of machines power the production of everything we need for our everyday lives, from food and medicine to fuel and electricity. Because machine downtime is very expensive, when things go wrong companies are seriously impacted.
To avoid such costly incidents, extend the life of equipment and reduce operational inefficiencies in the process, industrial plant operators are increasingly turning to predictive maintenance (PdM) solutions based on low power wireless technologies. Wireless PdM, or wireless condition-based maintenance, involves reliability engineers remotely monitoring data about the condition of industrial assets to catch defects and forecast problems before they escalate. In turn, the engineers can reduce maintenance to solely occasions necessary for the asset to continue operating at its full potential.
Research suggests it’s advisable, if not essential, for companies to prioritize machine reliability. According to the International Society of Automation, $647 billion is lost globally each year due to machine downtime. A 2017 ServiceMax study (conducted by GE Digital and titled After The Fall: Cost, Causes and Consequences of Unplanned Downtime) revealed 82 percent of companies experienced at least one unplanned downtime outage over the previous three-year period, while widely-reported 2012 research by analyst firm Aberdeen estimated that unplanned downtime can cost a company $250,000 per hour.
The wireless advantage
Plant maintenance teams at Louisville Gas & Electric, based in Kentucky, U.S., are tasked with monitoring the condition of boilers and piping to ensure equipment continues to operate safely and reliably, while also considering the possibility of failure. Previously, the continuous monitoring of these assets at the utility involved a costly, staff-intensive process of collecting and sending data to a main server stored at the plant. An individual engineer would then need to visit the plant and inspect the data on that specific computer. Determining issues and making maintenance recommendations could take hours.
Since the facility implemented Texas-based National Instruments’ (NI) PdM solution (incorporating NI Wireless Vibration Sensors with Bluetooth LE connectivity enabled by Nordic’s nRF52840 SoC, an NI gateway and NI InsightCM software) these same reliability engineers have been able to receive automated alerts and immediately access more data online than ever before. All they need is a device like a laptop, tablet or smartphone connected to the network, whether they’re working at an external office or another off-site location.
Typically, modern industrial plant applications involve the permanent installation of wireless vibration and/or temperature sensors on machinery to provide the comprehensive data needed for successful PdM. But thanks to robust and cost effective, low power wireless technologies like Bluetooth LE and the power of the IoT, many more wireless sensors can now be attached to many more machines throughout a large facility. These sensors then report critical data back to a central server or the Cloud, via a wireless gateway. When a nominated threshold for asset health is reached and an alert raised, reliability engineers have instant access to the machine’s condition data and analytics via asset management software, from anywhere.
Wireless networks also make it possible to monitor assets in hazardous environments and hard-to-reach locations, while drastically reducing the costs associated with the installation, maintenance and performance of the asset monitoring systems. As wireless sensors in industrial settings can’t always be easily recharged or replaced, it’s imperative they’re able to achieve long battery life – no mean feat given the power it takes to collect and send data up to hundreds of meters range and through many tangible obstacles. Wireless networks ensure the right balance is struck between throughput, range and low power consumption requirements.
The latest PdM solutions provide automated machine learning analytics through AI technology, allowing experts to spend even less time collecting data and even more time diagnosing and fixing issues. Essentially the IoT acts as AI’s eyes and ears, with human beings only needing to investigate once the AI has raised a red flag.
Imagine, for example, the logistical challenge engineers would face at Boeing’s Everett facility, the largest enclosed industrial space in the world, if they were forced to manually check the aircraft factory’s equipment. Fortunately, the aircraft manufacturer has developed and trained machine learning algorithms able to identify real-time data patterns and make accurate recommendations to its engineers within a few minutes. Boeing’s predictive maintenance tools have even helped its customers “reduce 80 per cent of maintenance burdens on a problem,” according to Dawen Nozdryn-Plotnicki, the company’s Director of Advanced Analytics for Digital Aviation and Analytics.
The predictive maintenance revolution is being driven by a combination of wireless connectivity, affordable wireless sensors, the IoT and AI technologies. Not only is more sensor data now available to provide superior analytics, but by bypassing the prohibitive expense of wired monitoring systems—providing engineers with greater problem-solving flexibility and improving overall equipment reliability—a company can achieve substantial savings on its bottom line. It’s therefore safe to predict wireless PdM is the way of the future for industrial plants of all sizes.