With metalworking companies across all industries facing increased demands, engineers need to rethink plant and maintenance tasks and operations. Customers are expecting greater manufacturing flexibility, more reliable, fast-turnaround product shipments and higher quality standards — all at lower costs.
To stay competitive, metalworking companies need to identify more efficient approaches to a variety of production tasks. This need for ongoing efficiency is leading metalworking companies to what some have described as the 4th industrial revolution. At the heart of this latest revolution is a greater reliance on smart, data-driven technologies.
Nowhere has data-driven technology taken a more prominent role as it has in predictive maintenance.
What is predictive maintenance?
Predictive maintenance (sometimes referred to as condition-based maintenance) is a failure inspection strategy that uses real-time data and models to predict when equipment components will wear down or malfunction so proactive corrective actions can be planned. It covers a wide array of topics, including failure prediction, failure diagnosis, recommended mitigation and suggested maintenance actions to be taken after failure.
In comparison to its predecessor strategies of reactive maintenance and preventative maintenance, predictive maintenance represents a more forward-thinking, cost-effective approach. With a reactive maintenance strategy, assets keep running until they fail. The problem is that such untimely failures result in unexpected and extended downtime and maintenance. With preventative maintenance, in contrast, problems are prevented before they occur. While this approach reduces or eliminates unplanned equipment failures and maintenance downtime can be planned, it does not allow for full utilization of an asset’s service life.
What are the benefits of predictive maintenance?
Predictive maintenance allows for planned downtime while avoiding premature maintenance so plants can get the full value from components. Predictive maintenance analyzes data gathered from numerous equipment sensors to provide a holistic view of asset health.
The concept behind predictive maintenance is not new. It has existed in some form for decades and is already a dominant strategy in many other industries. It has only been recently, however, that the technology, including the sensors, computers and software, has caught up and become affordable for more widespread use in a wide array of industries, including metalworking.
The data captured from the various connected systems and sensors provides unprecedented insights into the health and total service life of critical components, making smarter maintenance strategies possible. Artificial intelligence (AI) uses algorithms to find patterns and predict mechanical failures before they occur. Instead of simply scheduling maintenance procedures based on preset intervals, a machine learning system can analyze thousands of data points to prioritize maintenance and reduce failure risks.
Predictive maintenance remains one of the most effective approaches, to date, for equipment-maintenance scheduling. It has proven to offer significant bottom-line benefits by increasing equipment availability (productivity), reducing total maintenance costs and avoiding more costly repairs. By preventing major malfunctions, a predictive maintenance program also can help to reduce accidents in the plant to create a safer working environment.
What is monitored as part of a predictive maintenance program?
Machinery connected through an Internet of Things (IoT) platform sends continuous updates on a wide array of factors that affect long-term equipment performance. Some of the more commonly used condition monitoring techniques include:
New innovations in predictive maintenance are allowing for the monitoring of multiple factors simultaneously. For example, Parker’s SensoNODE Bluetooth-powered sensors catch performance fluctuations across a wide array of components and transmit the data through the Voice of the Machine Mobile App, which records data in real-time while also tracking historical performance.
In addition, the Parker Tracking System supports any predictive or diagnostic maintenance program.
Some metalworking plants are taking their predictive maintenance programs to new levels by adding machine-learning software. Although machine learning has been researched for years, its use in applying AI in industrial plants is now advancing more quickly.
What other technologies are making a difference in metalworking?
Automation technologies are also integral components of the 4th industrial revolution, even though metalworking has traditionally been slower than other sectors in its adoption of automated processes.
Robots and their smaller cobot counterparts have, more recently, been making their way into the metalworking industry. Now with lower costs, robots and cobots are seen as attractive solutions to counter skilled labor shortages.
Also gaining popularity are 3D printers, which have proven valuable due to their ability to manufacture more complex and lightweight parts and offer the design flexibility that’s necessary to compete in today’s highly competitive markets. Today’s 3D printers can create intricate parts directly from a CAD drawing, effectively eliminating the need for multiple lathe and mill setups. Until recently, 3D printing was almost exclusively used in the plastics industry, but it is now being adapted for metalworking.
How is technology shaping the future of the metalworking industry?
A new generation of technology is allowing metalworking plants to become more efficient and competitive.
Predictive maintenance innovations that can help reduce production downtime and contribute heartily to corporate bottom lines are key. Although many people in the industry may not think of the maintenance shop as being a prime location for investing in cutting-edge technology, the reality is that new maintenance metalworking technologies offer tremendous potential.
Article contributed by the IoT and the Fluid and Gas Handling Teams.
18 May 2021
18 Nov 2020
15 Oct 2020
15 Sep 2020
12 Aug 2020
18 Jun 2020