Don't miss MaximoWorld 2024, the premier conference on AI for asset management!

Experience the future of asset management with cutting-edge AI at MaximoWorld 2024.

Sign Up

Please use your business email address if applicable

AI for Reliability and Asset Management

AI for Reliability and Asset Management

The maintenance sector has experienced a profound shift in recent years. Labor force changes, more consumer demand, and supply chain issues have created a high-pressure situation. Maintenance managers must achieve better results with fewer resources. Fortunately, technology can help. Exciting new tools are delivering solutions that meet the reality of the modern workforce. Artificial intelligence (AI) is at the forefront of this new tech revolution and is the tool with the most promise for transformation.

How AI Can Transform Maintenance Operations

Artificial intelligence represents the next step forward on the path toward continuous improvement. New technology can greatly improve productivity, reduce downtime, and help maintenance teams meet demand by scaling operations. This article explains what AI can do and how to use it in maintenance operations.

But first, let’s just get one thing out of the way. There are, understandably, some concerns about the possible impact of AI on the labor force. To be clear, AI is not going to take away jobs. There will always be a need for human workers at all skill levels.

AI helps maintenance teams work better and faster. It is a tool that maintenance teams can use to improve their work and make it easier. It is just one part of a bigger process. Now, let’s get into what that means in practice.

"AI helps maintenance teams work better and faster."

AI: The Right Tool at the Right Time

AI didn’t appear out of nowhere in the maintenance sector. If anything, AI is the answer to some of the problems that have been plaguing maintenance teams for years.

“AI is the answer to some of the problems that have been plaguing maintenance teams for years.”

Over the past several decades, industrial operations have increasingly adopted Internet of Things (IoT) and predictive maintenance technologies, ramping up their capabilities. IoT sensors are cheaper now and cloud technology makes it easier to stream, store and analyze condition monitoring data. Organizations can now easily monitor the condition of more assets using things like vibration and temperature data.

Other advancements are more sensors connected to assets and parts. These sensors gather data for monitoring and the software instantly receives and stores the data for analysis.

That’s the heart of any predictive maintenance program: collecting data on key assets and then monitoring that data for signs of a new or developing machine fault. Changes in an asset’s vibration or temperature levels can be early indicators of something going wrong. The sooner maintenance teams know about a problem, the more easily they can correct it.

These are all good things. More data helps predict when repairs are needed and schedule them at convenient times. The more data points you have, the more effectively you can manage your resources and supervise multiple worksites.

However, handling all this data is a significant challenge for human workers. It is simply not possible for a team of human personnel – no matter how skilled – to sift through the mountains of condition monitoring data and make sense of it. As operations grow, adding more assets and more worksites, the amount of data coming in also grows exponentially.

That’s where AI comes in. AI’s speed, automation and pattern-finding capabilities can put data to work faster than ever before, transforming maintenance operations. AI gives maintenance teams the tools they need to actually use the data they’re collecting.

AI and Proactive Maintenance

Predictive maintenance harnesses condition monitoring data and uses that data to make predictions about when each asset will need repairs, or a belt change, or other maintenance work. Done right, predictive maintenance saves time and reduces costs. It also means machines can stay up and run longer since maintenance teams can fix problems before they get serious enough to take machines offline.

The problem is that reliability engineers can’t analyze all the data they receive. AI technology can put that data to use. AI can read and analyze large volumes of data at an enormous speed, far faster than a human. The tools don’t get tired or careless, reducing the chances of human error.

AI tools help technicians quickly identify and fix problems by spotting data anomalies, making predictive maintenance easier and faster. Technicians receive automatic notifications when issues arise, letting them inspect and repair assets efficiently.

Maintenance teams can utilize AI to analyze vibration and temperature data. AI can notify technicians of any changes that may indicate a potential issue. These changes may include an unbalanced rotor or a misaligned shaft.

AI tools automate monitoring of large operations, without burdening your team or hiring more people. AI also helps maintenance teams monitor remote locations, such as oil rigs or plants, effectively and without missing important information.

AI Diagnostics

Additionally, AI can go a few steps beyond simply detecting anomalies. Some AI algorithms are already capable of diagnosing the root causes behind those anomalies. That means AI can look at the data and figure out whether the changes in vibration levels are due to a bearing defect, for example, or a misaligned shaft.

From there, managers make informed decisions about scheduling repairs. Is this a minor issue that won’t impact the asset’s productivity? Will this problem eventually cause a shutdown if it’s not addressed? An AI tool like this can also help maintenance teams manage their spare parts inventory and stay in compliance with the relevant regulations.

In some cases, these AI tools can even guide the technician through the necessary steps to correct the issue. This ability to diagnose and make informed suggestions is known as prescriptive maintenance.

How AI Identifies Potential Machine Faults

AI can be a powerful tool to help maintenance teams make smart, data-driven decisions. AI algorithms analyze the available data, consider a range of possible outcomes, and make recommendations based on those potential outcomes.

In a way, this is similar to what technicians have been doing for generations. Traditionally, vibration specialists have gathered condition monitoring data about assets – even when that meant just listening to a motor’s unique rattle and noticing that something was off. They use this data to diagnose machine faults and decide whether those faults are serious enough to address right away.

Today, this kind of human expertise is still invaluable. But as work sites have grown more complex, maintenance teams simply cannot keep everything up and running on their own. Not only has the equipment itself grown more complex, but the average plant has more assets than ever before.

Managers oversee many worksites and cannot manually monitor all critical assets due to limited resources. To keep up, forward-thinking maintenance teams are already starting to use AI-powered condition monitoring tools to help augment their capabilities and assess condition monitoring data at scale.

Setting Thresholds and Detecting Anomalies

AI can learn to recognize anomalies in condition monitoring data through various training methods. The most obvious involves manually setting thresholds. When vibration levels cross that threshold, the algorithm recognizes this and automatically issues an alert.

But AI can do more than just enforce a predefined threshold. AI can learn on the job, so to speak. The algorithm improves as it takes in more data and studies more outcomes. In other words, the longer AI operates, the more accurately it can identify truly problematic anomalies and issue timely alerts and notifications.

AI also excels at pattern recognition, and it is constantly refining the patterns that it creates. The algorithm is self-correcting and can fine-tune the thresholds for condition monitoring data by comparing data to results. This means AI can “learn” the specific needs of each asset in an operation. For example, suppose one pump has a different vibration signature from the others in the same facility, maybe because of the piping or some other harmless reason. AI can learn to account for the difference in that pump’s vibration signature. In the same way, AI can learn which assets need to be lubricated more frequently or need a part change more often because of the environment they’re operating in.

AI and Prescriptive Maintenance

AI’s pattern recognition is what allows it to diagnose machine faults. Again, the more data AI takes in, the more effectively it can diagnose the root causes behind changes in condition monitoring data. The algorithm continues to fine-tune its diagnostic capabilities over time. It also learns by studying work order data and the response time of human technicians.

But there’s another tool at play here: generative AI. Generative AI allows AI to communicate effectively with human maintenance teams. While prescriptive maintenance refers to AI’s ability to diagnose faults and give very precise guidance on correcting those faults, generative AI is what drives that communication so AI can create clear instructions for maintenance crews to follow.

The combination of diagnostic capabilities and communication are what can potentially make AI a game changer.

Phasing in Industrial AI Implementation

AI has the potential to transform just about any operation, but at the same time, implementing AI is no different than implementing any other new strategy. It’s a serious undertaking and should be treated with care.

It’s a good idea to start with a pilot program. Pick a small group of assets to begin testing the program. Make sure there are well-defined benchmarks in place and that everyone on the team has a shared understanding of the goals.

Along the way, perform tests regularly to ensure the new technology is working the way it should. It’s also important to check in regularly with employees to ensure they are comfortable working with the new tools.

Once the pilot program is over, assess what’s worked and begin expanding the program.

Looking Ahead

AI technology is still full of untapped potential, so you can expect to see dramatic new developments in the months and years to come. At this point, the tool’s capabilities and potential are still being explored, but there’s great promise for the future.

What’s the best way to take advantage of AI? Get started now. Teams that have already shifted to a predictive maintenance approach – adopting tools like wireless sensors and making data-driven decisions – are already well-positioned to start working with AI. Facilities that haven’t yet moved to predictive maintenance would be well-advised to do so now.

“Teams that have already shifted to a predictive maintenance approach – adopting tools like wireless sensors and making data-driven decisions – are already well-positioned to start working with AI.”

It’s also a good idea to consult with experts in the field. Some operations already have AI and condition monitoring experts in-house. Others may want to partner with outside experts on a onetime or ongoing basis. Tapping into expertise means being able to get the full benefit out of the new technology while avoiding pitfalls and accidents.

The full benefits of AI are clear: the technology allows industrial operations to scale and expand their offerings while keeping costs low and making the most out of limited resources. Ultimately, AI moves teams closer to the goal of continuous improvement and achieving greater productivity, uptime and reliability.

Aaron Merkin is Chief Technology Officer of Fluke Reliability, home to eMaint, Pruftechnik and Fluke Connect. His responsibilities include developing and executing IIoT strategy and leading the technology team in the continued creation of innovative solutions for customers. He has more than two decades of experience developing enterprise software across a variety of industries and markets. Merkin holds a master’s degree in computer science and a bachelor’s degree in mathematics.

Aaron Merkin

Aaron Merkin is Chief Technology Officer of Fluke Reliability, home to eMaint, Pruftechnik and Fluke Connect. His responsibilities include developing and executing IIoT strategy and leading the technology team in the continued creation of innovative solutions for customers. He has more than two decades of experience developing enterprise software across a variety of industries and markets. Merkin holds a master’s degree in computer science and a bachelor’s degree in mathematics.

ChatGPT with
ReliabilityWeb:
Find Your Answers Fast
Start