Trelleborg: Cognitive sealing a safer, more cost-effective approach

TRELLEBORG, Sweden—With the onset of new technology comes new terminology, and that has never been more evident as Trelleborg A.B. advances its work in the digitalization of sealing systems.

Since 2016, the global seal and gasket producer has endeavored to improve both the safety margin and the bottom line for its customers—whether that customer is an offshore oil company or a producer of wind turbines—through "cognitive sealing," which, in turn, allows for "predictive maintenance" of sealing systems.

A third relatively new term makes cognitive sealing and predictive maintenance possible: the Internet of Things, a general term for the interconnection between computing devices and mechanical and digital machines, and the ability to transfer data that is gleaned between those elements over a network without requiring human-to-human or human-to-computer interaction.

"The Internet of Things is where physical and digital objects get connected," said Johannes Kunze von Bischhoffshausen, manager for digital transformation and Internet of Things at Trelleborg Sealing Solutions. "More customers are beginning to monitor their assets remotely, and this includes any assets in the field, wind power or otherwise. Trelleborg is applying this same principle to its 30 manufacturing sites around the world, taking it to the next level and moving into the predictive area."

According to von Bischhoffshausen, predictive maintenance means not simply looking at what a machine is doing in a practical, literal sense, but also predicting with surgical accuracy what a specific mechanical system is going to do—and more importantly, when the machine might break down or the seal might fail.

The safety and savings brought by such knowledge, while not completely quantifiable yet for a company's bottom line, almost certainly brings value to a customer because seals, gaskets and dampers often can be the weakest part of an operation with the greatest consequence for failure.

"The cognitive sealing approach focuses on exactly this challenge," von Bischhoffshausen said. "Customers have been asking: 'Can you place a sensor inside a seal? Can you predict when the seal will fail?' There are different ways to achieve this. While we don't place a sensor inside a seal, we combine data from sensors all around a sealing system—from vibration to temperature and pressure data. This cognitive approach allows us to infer seal health indirectly.

"Customers see this value on a more strategic level. They may not be able to calculate the savings from downtime today, but they know they have to move toward digitalization to gain a competitive advantage."

How a cognitive approach works
More than ever, companies are seeking to increase the safety of their machines and applications, while at the same time aiming to reduce maintenance and the cost of unexpected downtime.

In an aircraft, more than 6,000 sensors capture more than a terabyte of data in a single flight. It's information that is then used for predictive maintenance to prevent future stresses, cracks and failures, for both sealing and other mechanical systems.

Such knowledge obviously is crucial to the aviation industry, and is no less so to the operators of an offshore oil rig, wind turbine or injection molding machine.

Von Bischhoffshausen brings his expertise in the Internet of Things to Trelleborg Sealing Solutions from the IBM Watson Division, where he helped develop "actionable insights" based on data gleaned from this cognitive, predictive approach.

"The challenge is, what do you want to do with the information?" von Bischhoffshausen said. "And this is quite a huge challenge."

Interpreting the data gleaned from pressure, temperature and vibration sensors within a sealing system is done by creating algorithms that are unique to each customer's processes.

Trelleborg has developed a new IoT lab for this cognitive sealing concept, the company said, which prototypes electronics and sensors, and contains an IoT test rig and demonstration area. The idea, von Bischhoffshausen said, is to evaluate IoT technologies and build prototypes in a short timeframe.

The ability to do this is based on innovative sensor technology, edge computing, artificial intelligence and Trelleborg's expertise in materials and geometries for better performing seals, he said.

While its predictive maintenance work with customers is a proprietary process for Trelleborg, the real secret is the expertise that Trelleborg has gained through its success and failures, von Bischhoffshausen said.

"We work with selected customers," he said. "What type of sensors are needed? Does a customer just need the sensors or both the sensors and IT support? Is the customer an OEM or an aftermarket parts supplier? Is predictive maintenance even feasible for what the customer is asking? There is a lot behind it all."

As it is in aviation and aerospace, seal performance can be critical to safety in the hydraulic wind power plant industry.

In a storm, wind turbine blades (their pitch controlled by a hydraulic sealing system) move to a neutral position so that they do not over-spin.

"As mentioned before, with an offshore oil rig or wind turbines, really anything with a high cost of downtime and maintenance, this cognitive approach adds value. Seal failures in these industries were causing catastrophic damage, a complete breakdown," von Bischhoffshausen said. "It is in this regard that remote monitoring of a seal component is safety-related. And we are finding that more and more industrial equipment is automated in this way. Today we are not seeing people sitting at an actual machine to supervise and operate it. This encourages an even more accurate predictive functioning modality."

A look back at a novel approach
Trelleborg Sealing Solutions began to study the overall idea of cognitive sealing in 2016, von Bischhoffshausen said, when the company became aware that digitalization of sealing systems would be "highly relevant."

"Trelleborg was always at the forefront, developing an online catalog and digital tools and applications," he said. "This was already a journey we had taken, building a huge work force that could work on such projects with IT people, engineers and data scientists. We started with internal testing to see if these theoretical concepts would even work, then we worked with certain pilot customers to see if it could be applied to them."

Trelleborg and von Bischhoffshausen have reduced significantly the development time for customer projects today, though it was a big learning curve in the beginning.

"We learned from our failures. We tried new things and we learned from them," he said, citing an offshore project that was a "sprint" for Trelleborg, as the company had a solution for the customer within six weeks.

"We can turn things around quickly now," von Bischhoffshausen said.

While this cognitive, predictive approach to seal maintenance can offer value, customers still turn to Trelleborg for well-engineered seals that avoid leakage.

"Trelleborg seals are a high quality, and often need no predictive maintenance at all," von Bischhoffshausen said. "In the end, it is still about the empirical nature of the seal."

And cognitive sealing and seal production are not mutually exclusive. Data points, whether they are used for predictive maintenance or manufacturing, still represent crucial information that can be used to improve the seal composition itself.

"Although this was not the first intention of cognitive sealing, we can start to think about the next step," von Bischhoffshausen said. "OEMs typically do this, where they will look at the data points to improve the overall application and performance of the seal. What we learn from the field will be part of our focus for the next years."

As new technologies like cognitive sealing arrive as part of digitalization, von Bischhoffshausen said Trelleborg continues to look over the horizon at other blossoming applied sciences.

"Besides digitalization, we are looking at two other big areas on our radar: sustainability and electrification," he said. "In sustainability we continue to look for ways to reduce our footprint in our processes and our products. This is a big strategic priority for us.

"With electrification going into more and more vehicles in e-mobility, we will see new sealing technologies used for the motor and batteries."

And what will those new sealing technologies be called? Like the world of digitalization, sustainability and e-mobility likely will add more unique terms such as cognitive sealing, predictive maintenance and the Internet of Things to industry lexicons for companies like Trelleborg.