Predictive maintenance proving to be a successful AI use case

John P. Desmond, AI Trends editor

More companies are successfully using predictive maintenance systems that combine AI and IoT sensors to collect data that predicts breakdowns and recommends preventive actions before a breakdown or machine failure, with an AI use case showing proven value.

This growth is reflected in the optimistic forecasts of the market. The predictive maintenance market is worth $6.9 billion today and is projected to grow to $28.2 billion by 2026, according to the report. IoT Analytics: Hamburg, Germany. The company has more than 280 vendors offering solutions in the market today, with plans to grow to 500 by 2026.

Fernando Bruegge, Analyst, IoT Analytics, Hamburg, Germany

“This research is a wake-up call for those who argue that the IoT is failing,” said analyst Fernando Breguet, author of the report, adding: “For companies that own industrial assets or sell equipment, now is the time to invest in predictive maintenance. – typical solutions”. And, “Enterprise technology companies should prepare to integrate predictive maintenance solutions into their offerings,” suggested Bregge.

Here’s an overview of some experience with predictive maintenance systems that combine AI and IoT sensors.

Manufacturer of aircraft engines Rolls Royce is deploying predictive analytics to help reduce the amount of carbon produced by its engines, as well as optimizing maintenance to help customers keep planes in the air longer, according to a recent report. CIO:.

Rolls-Royce has built an Intelligent Engine platform to control engine flight by gathering data about weather conditions and pilot flight. Machine learning is applied to the data to customize maintenance routines for individual engines.

Stuart Hughes, Chief Information and Digital Officer, Rolls-Royce

“We tailor our maintenance routines to make sure we’re optimizing the life of the engine, not the service it says in the manual,” said Stuart Hughes, chief information and digital officer at Rolls-Royce. “It’s really variable maintenance, treating each engine as an individual engine.”

Customers experience fewer service interruptions. “Rolls-Royce has been monitoring engines and charging hourly for at least 20 years,” Hughes said. “That part of the business is not new. But as we’ve evolved, we’ve come to treat the engine as a unique engine. It’s much more about personalizing that engine.”

Predictive analytics is used in healthcare as well as manufacturing industries. Kaiser Permanente, an integrated managed care consortium based in Oakland, California, uses predictive analytics to identify non-intensive care unit (ICU) patients at risk for rapid deterioration.

Although non-ICU patients requiring unexpected ICU transfer make up less than 4% of the total hospital population, they account for 20% of hospital deaths, according to Dr. Gabriel Escobar, research division and regional director. , Hospital Operations Research, Kaiser Permanente Northern California.

Kaiser Permanente Practicing Predictive Maintenance in Healthcare

Kaiser Permanente developed the Advanced Alert Monitor (AAM) system using three predictive analytics models to analyze more than 70 factors in a given patient’s electronic health record to create a composite risk score.

“The AAM system synthesizes and analyzes vital statistics, laboratory results and other variables to generate hourly deterioration risk scores for adult hospital patients in medical-surgical and transitional care units,” said Dick Daniels, executive vice president and CIO of Kaiser Permanente. The CIO account. “Remote hospital teams assess risk scores every hour and notify hospital rapid response teams when possible deterioration is detected. The rapid response team conducts a bedside assessment of the patient and coordinates the course of treatment with the hospital.”

In her advice to other practitioners, Daniels recommended focusing on how the tool will fit into the workflow of healthcare teams. “It took us about five years to do the initial mapping of the electronic medical records and develop the predictive models,” Daniels said. “Then it took us another two to three years to turn these models into a live web services application that could be used operationally.”

In the food industry as an example, the PepsiCo Frito-Lay plant in Fayetteville, Tenn., successfully uses predictive maintenance with 0.75% equipment uptime and 2.88% unplanned downtime, according to Carlos Calloway, site specialists. reliability engineering manager, account PlantServices:.

Examples of monitoring include: ultrasound-confirmed vibration readings helped prevent PC combustion blower motor failure and shut down the entire potato chip section; Infrared analysis of the main pole of the station’s GES automated warehouse identified a hot fuse box that helped avoid shutting down the entire warehouse; and elevated acidity levels were found in oil samples from the baked extruder gearbox, indicating oil degradation, which prevented the Cheetos Puffs from being shut down.

The Frito-Lay plant produces more than 150 million pounds of products annually, including Lays, Ruffles, Cheetos, Doritos, Fritos and Tostitos.

Types of monitoring include vibration analysis used in mechanical applications, developed with the help of a third-party company that sends alarms to the plant for investigation and resolution. Another service partner performs quarterly vibration monitoring on selected equipment. All engine control center rooms and electrical panels are monitored by quarterly infrared analysis, which is also used on electrical equipment, some rotating equipment and heat exchangers. Additionally, the plant has been doing ultrasound monitoring for more than 15 years, and it’s “like the pride and joy of our site from a forecasting perspective,” Calloway said.

The plan features a variety of products from UE Systems of Elmsford, NY, a supplier of ultrasound instruments, hardware and software, as well as training for predictive maintenance.

Louisiana Aluminum Plant Automation Bearing Service

Bearings, which wear over time in different weather and temperature conditions in automobiles, are prime candidates for IoT monitoring and predictive maintenance with AI. It Noranda Alumina The factory in La Gramercy is seeing great benefit from its investment in a system to improve bearing lubrication in its production equipment.

The system resulted in a 60% reduction in bearing changes in the second year of using the new lubrication system, which translates into savings of nearly $900,000 in bearings that did not need to be replaced and avoided failures.

“Four hours of downtime is about $1 million in lost production,” said Russell Goodwin, reliability engineer and millbuilder instructor at Noranda Alumina, in a PlantServices estimate based on presentations at the Leading Reliability 2021 event.

The Noranda Alumina plant is the only alumina plant operating in the United States. “If we close, you have to import it,” Goodwin said. The plant experiences widespread dust, dirt and caustic substances that complicate efforts to improve reliability and maintenance practices.

Noranda Alumina monitors all engines and transmissions 1,500 rpm and above with vibration readings, and most below 1,500 with ultrasound. Ultrasonic monitoring, sound beyond the limits of human hearing, has been introduced at the plant since Goodwin joined the company in 2019. Back then, fat monitoring had room for improvement. “If the grease wasn’t visibly coming out of the seal, the mechanic didn’t consider the rotation complete,” Goodwin said.

He noted that after the introduction of automation, the lubrication system has improved dramatically. The system was also able to detect bearings in a belt whose bearings were wearing too quickly due to contamination. “Tracking with the tools helped prove that it wasn’t improper lubrication, but rather, the bearing was made incorrectly,” Goodwin said.

Read source articles and information here IoT Analytics:, in CIO: and in PlantServices:.

Source link