237 - Bringing Ultrasound to iIoT with Blair Fraser

Rooted in Reliability: The Plant Performance Podcast

Bringing Ultrasound to iIoT with Blair Fraser

We’re glad to have Blair Fraser, the Global Director for IIoT Solutions at UE Systems. He’s been our guest on topics like Artificial Intelligence, IIoT, and other areas relating to tech and maintenance and reliability. Blair previously worked at Quartic and Lakeside Process Controls, getting involved in the tech side of reliability. His earlier career influences got him involved in maintaining assets on the plant floor. Blair will be giving us more insight on bringing ultrasound in the iIoT space.

Blair Fraser of Quartic.ai shares some important insights as well as:

• How do we bring Ultrasound to IIoT?
• How is Ultrasound being brought into IIoT?
• Is Ultrasound qualitative or quantitative?

…and so much more!

How do we bring Ultrasound to IIoT?

Vibration is fairly simple since we know what’s being measured and the parameters. But ultrasound isn’t always as quantitative. Sometimes, it’s qualitative. People say vibration is simple, but when you get into it, it’s not. They base that statement on the level of training people get to become, specifically CAT IV vibration people. Vibration is the more common technique to diagnose bearings since it’s a great technology. However, ultrasound fits in with vibration. Ultrasound, thermography, and oil analysis all fit into a bigger ecosystem. The main difference comes when you think of it from an IoT view of trying to capture data. But the goal should include insights as well. Data is the raw material to get to the new goal, which is insights. We’re trying to get those insights to make better decisions on the health and performance of an asset.

97% of our data doesn’t get used, yet we feel compelled to make more data. The challenge with any measuring technique is knowing what you’re trying to accomplish. First, we look at the IoT ecosystem. Blair likes to start by putting up all possible solutions out there. That gives the view that we shouldn’t have any plant issues since all problems have solutions. But the reality is that we still have issues, and not every problem has a solution. So, what are you trying to solve? That’s where ultrasound comes into IoT.

Ultrasound uses

Ultrasound is good at:

• Detecting or measuring friction
• Measuring impacting
• Measuring turbulence

These make ultrasound, especially Structure Borne Ultrasound, great for things like compressed air leaks, bearing friction, cavitation, and leaky valves. That’s because these need pinpoint solutions. But that’s not to say having an ultrasound means no more failures. But it has very specific applications within the IoT ecosystem.

How is Ultrasound being brought into IIoT?

It’s a little bit of a quantitative and a qualitative approach, with our human senses also playing a part. Some of the challenges present with bearings come with getting quantitative data on good and bad bearings. This gets addressed, specifically, through using ultrasound, which gets used to validate other technologies.

Rolling element bearings, also called anti-friction bearings, are meant to reduce friction. We’ve known this for many years. In studies, the question is usually, “How do most rolling element bearings fail?” Studies show that over 80% prematurely fail because of lubrication issues. This is interesting since we already know the issue is lubrication, but there’s no solution to solve it.

Lubrication is our superhero against friction. But that’s what we’re fighting against when doing bearings and lubrication. There’s time-based lubrication that uses advanced calculations. In his early career days, Blair used to blindly plug in the numbers into a calculator. But looking from an IoT point of view, you need to check the variables. You have speed which impacts the friction and wears the grease or lubricant in a bearing. But nowadays, we’re moving to more energy conservation by controlling flow pressure with variable speed drives. The speed here is changing, including on and off duty cycles. Other coefficients are also present, like the room’s humidity, temperature, and vibration. But often, that doesn’t reflect what’s happening in the plant.

People think that just because there’s data in the cloud that it’s a digital twin. A digital twin should be reflective of the current operational and health state of that equipment. If you can have such a digital twin then feed it into that calculation, it would be great. But we don’t. So, we use ultrasound to measure the friction. We start to think about what we can do with friction. If the friction goes down then it tells us we need lubrication.

Is Ultrasound qualitative or quantitative?

We know that if you can set a baseline on that bearing, the Structure Borne Ultrasound will measure the change in that friction. We’ll know the condition of that bearing when we see an eight-decibel change. You can also listen as the lube goes in, helping you judge based on the sound. The decibel change will indicate that the thin film on the bearings has gone away and it needs replenishing. We haven’t used time or any calculations for that, just the condition. It’s a quantitative number that can be put into the system to say a bearing needs lubrication.

We need insights

When you get an alert from an ultrasound system saying a bearing needs lubrication, you know exactly what to do. That’s a very descriptive quantitative alert but what we need is prescriptive. So, a bearing needs lubrication. What next? You can lube it or take a step further to determine how much lube is necessary. So, when looking at the 80% of premature bearing failures caused by lubrication, nearly 15% of that gets caused by inadequate lubrication. That could be either over or under greasing, with the former being more common.

Quantitative ultrasound values come in decibels. So, for someone unfamiliar with ultrasound, you let them know they’re measuring friction. Dispense grease into that bearing, which will reduce the friction. The lower the friction the lower the decibel until you reach the baseline. That answers two questions quantitatively:

1. When does a bearing require grease?
2. How much grease is required?

That’s in an ideal situation. In most cases, bearings may have defects that stop you from getting down to baseline. For such challenges, some logic has to come in from humans or automated systems. You should consider more than that best-case scenario.

UE Systems’ OnTrak

UE Systems’ IoT platform is OnTrak, with an exciting new feature coming out as well. But what does it provide for organizations and what are its applications apart from measuring friction?

We’ve mentioned things like bearing monitoring or lubrication health monitoring, leaky valves and cavitation in pumps as the typical applications of Structure Borne Ultrasound. Then there’s the Airborne Ultrasound which gets primarily used for leak detection, the detection of corona discharge, partial discharge, and things like that in electrical distribution systems. The OnTrak can use either Structure Borne or Airborne Ultrasound to find the value.

Customers and end-users don’t need IoT. They need a solution to their problem. Whether IoT can help solve that problem is like a piece in a puzzle. So, with the OnTrak, you have specific use case applications for bearing health and lubrication monitoring.

Nearly everyone now is measuring vibration. But if you look at the P-F curve, where are people spending their time and money? The predictive domain, which comes after the failure has started to occur. So, would you rather prevent a failure or detect it as early as possible? Everyone’s racing to have systems that detect failures first rather than preventing them.

So, the OnTrak system focuses on reducing those bearing failures from happening. It’s also a pro-active lubrication needs system to tell you when and how much lubrication to provide to that bearing.

Failure prevention is better

In reliability, we need to start focusing more on preventing failures than detecting them after they happen. That’s whether it’s in the design phase, spare parts storage, proper lubrication, or even installation. After all, predictive maintenance is waiting for something to fail before we react. We need to start celebrating preventing failure even though it’s not as quantitative as dealing with a failure.

With the OnTrak, we looked into how people were using our ultrasound technology. The data got collected every month by our people. It was to determine if a bearing needs lubrication and, if so, how much. It also helped in picking up defects that go beyond the lubrication failure mode. With the current pandemic, we needed a way to get the data without sending people to go fetch it. So, the OnTrak gives people access to this data anywhere, any time, and from any device in real-time.

Another perk with this system is that it works with other systems out there like OSI PI or Rockwell. Since it’s your data, you can send it wherever you want. There’s a cloud platform designed to work with the OnTrak system called UE Insights designed to measure friction and ultrasound. This system gets used to monitor lubrication as well as to send data to a larger system for a more comprehensive overview. It meets the key need that IoT aims for, which is to send data to multiple places.

What’s included in UE Insights?

AI is going to be a gamechanger to sift through all the data for insights you couldn’t find before. Now we need to look into where we’ll use AI within ultrasound. We’ve found that we don’t need advanced calculations or AI to make insightful decisions on the life and health of our bearings. Then it’ll come into play when you integrate ultrasound data into other technologies. That will use AI and ML to make things easier with the UE Insights platform.

The biggest value in ML and AI will come when the data leaves our OnTrak system on UE Insights to go into a larger system. There, you’ll have access to:

• Process data
• Ultrasound
• Vibration
• Work order history

Using OnTrak

One of the challenges that came up with this system is that we can’t keep up with demand at the moment. However, if everyone goes to the study stating 80% of bearing issues relate to lubrication, they’ll find the most common issues. These include insufficient lubrication quantity, using the right grease, and contamination. Even with all the information out there, human error still comes in.

So, unfortunately, the OnTrak doesn’t eliminate those failure modes of contamination and using the right grease. That’s where an automatic lubricator comes in. These get installed on a bearing and get based on time, providing the right grease type at the right time. So, we were able to integrate the technology of the ultrasound-guided or condition-based lubrication with the convenience and safety of the automatic lubrication devices. That’s the update getting released.

Automatic lubrication devices were still based on time, with some having vibration centers built in. But when it dispenses, it tries to ensure the equipment is running and influence the duty. These are still on theoretical calculations but dispense with really good accuracy. It also remains contaminant-free and that you’re using the right grease.

We’ve formed partnerships with these automatic lubrication devices, but they’re not going to be based on time. We get the value of the friction coming in from the OnTrak. It senses the change and dispenses, regardless of the ambient temperature and vibration, or how much time has passed. It still happens on condition anywhere, anytime, through a phone. You can lubricate that bearing with precision through ultrasound-guided feedback without having to physically show up. That’s all thanks to the alerts you get.

One customer that tested this system found a 95% saving in lubrication tasks by not having to physically attend to the bearing. That’s because it even does an automatic PM on the bearing, with an alarm to notify you of the different reorder levels. You can get it on UEsystems.com/smartlube.

More from UE Systems

The pandemic has created a lot of restrictions. So, to ensure people can still learn new skills, we developed an online ultrasound training. We have a forecast system for very critical bearings where you can record the friction level, the sound recordings, and the WAV file. From that, you can do spectrum analysis with it. We then took that technology and brought it to electrical assets as part of our foresight systems. You can now monitor electrical cabinets, distribution cabinets, MCC panels looking for corona-partial discharge all online. We came up with the OnTrak, the UE Insights, and now the release of the SmartLube technology.

For those interested in ultrasound technology, we’ll have releases happening from early 2021 to change the game further. Just as we’re looking to change things in the lubrication area, how else is this technology being used:

• Steam trap monitoring
• Valve leak monitoring
• Cavitation

Eruditio Links:

• Eruditio
• HP Reliability
• James Kovacevic's LinkedIn
• Reliability Report

Blair Fraser Links:

• Blair Fraser LinkedIn
• UE Systems
• UE Systems 4Cast
• UE Systems 4Site
• UE Systems OnTrak
• UE Systems SmartLube

Download RSS iTunesStitcher

Rooted In Reliability podcast is a proud member of Reliability.fm network. We encourage you to please rate and review this podcast on iTunes and Stitcher. It ensures the podcast stays relevant and is easy to find by like-minded professionals. It is only with your ratings and reviews that the Rooted In Reliability podcast can continue to grow. Thank you for providing the small but critical support for the Rooted In Reliability podcast!

The post 237 – Bringing Ultrasound to iIoT with Blair Fraser appeared first on Accendo Reliability.