BY: Mike Chalmers
This article originally appeared in the January/February 2024 Issue of Wire Rope Exchange.
Around the globe, the standard best practice to determine whether a synthetic rope is safe for use is manual visual inspection. Basically, an inspector – or in many cases the operator, deckhand, or general worker most closely at hand – looks at a rope and decides if it is safe to use, or not.
This creates incalculable risk because that process is at root subjective and qualitative – there is no quantitative metric that expresses the quality of a rope that can be derived from a person’s visual inspection. And so, there is no tool to measure quality consistently across ropes, and no means to analyze line health across a vessel or fleet.
Enter Scope—a team of engineers, designers, and rope experts who have developed a product that is revolutionizing the long-standing inspection process through applied deep learning.
Scope has built and put into use automated inspection systems for critical synthetic rope applications, utilizing deep learning technology to identify and assess line wear and damage. Their systems enable operators to take data-driven actions to repair or replace failing lines before catastrophic incidents occur. And this is all accomplished in real-time, with a level of accuracy exceeding that of traditional visual inspection, with software that is continuously learning.
We’ll be delving into an application that requires inspecting miles of rope each time it is used, but before we get to that, let’s meet the team and learn more about the tool.
A Blend of Skills
Behind Scope are co-founders Justin McCoy, CEO, who specializes in Artificial Intelligence and Machine Learning (AI/ML) product design and user experience; Mike Poroo, COO, who specializes in in rope systems engineering; and John Hohm, CTO, who specializes in software engineering and Development/Security/Operations (or “DevSecOps”, defined by Amazon as “the practice of integrating security testing at every stage of the software development process”).
All three have extensive experience commensurate with their roles at Scope and the game-changing technology they’ve crafted. McCoy designed WIDOW (Web-Based Information DOminant Warfare)—the official mission planning tool of the U.S. Air Force—and led teams with the Department of Defense and U.S. Air Force Special Operations Command in the design of MAIZE, an online platform designed to counter disinformation campaigns with applied artificial intelligence.
Poroo is a third-generation owner of SWOS—an industry leader in the engineering and fabrication of rope systems. He’s also on the board of directors of the Associated Wire Rope Fabricators (AWRF), is a member of the Cordage Institute, and has led multifaceted teams developing innovative fiber rope solutions for SpaceX, the U.S. government, and several major energy producers.
Similar to McCoy, Hohm served on the WIDOW project, as lead engineer. He also led engineering teams with the Department of Defense and U.S. Air Force Special Operations Command building intelligent systems at scale, as well as the U.S. Navy, drafting interference mapping applications for 5G radar systems.
Essentially, the three have maximized their shared expertise and harnessed the power of AI to administer safer, smarter, and more efficient synthetic rope inspections. Their initial product, Scope Control V1 is a standalone inspection unit designed specifically for placement between two stringing units as line is spooled from one machine to the other. Once inside the unit, multiple cameras surround the subject rope, capturing a 360-degree view of the line.
Teaching the Machine to Learn
At the core of Scope’s system is the Insight Engine, a deep-learning neural network that has been rigorously trained to identify defects in rope with exceptional accuracy. Through extensive residual break strength testing and training, the Insight Engine can detect anomalies with exceptional accuracy, and can even predict Residual Break Strength (RBS) within +/-5 percent of a line’s actual RBS. And as more data is acquired through inspection scans, the machine learning platform continuously improves its performance.
Ultimately, the Insight Engine synthesizes information from each scan into a user-friendly graph, displaying the health of each individual line segment in real-time. The data it measures the results against is based on thousands of rope break tests that the team performed in order to train the system’s algorithms to understand how the degree of abrasion observed translates back to rope strength. The outcome allows operators to receive immediate feedback on the scanned line segments, facilitating quick and informed decision-making.
Additionally, the system precisely tracks pull speed and position on the line, making it simple to pinpoint the location of any damage. With every scan, the Scope system generates valuable data points that are securely stored online. This enables operators to review the line and formulate a repair plan if needed, while fleet managers can immediately access the scans and relevant images. In the event of a repair, the system will “remember” the position of the anomaly and slow the operator down for a closer look on reinspection of the line.
And it delivers real-time reporting and notifications, removing the need to manually create an inspection report—streamlining the inspection process for everyone involved, and making universally consistent data readily available.
With that data in hand, technicians can spend their time addressing root issues instead of looking for them. And they are likely to have captured signs of wear and tear, damage, or defects in synthetic ropes that may not be easily visible to the human eye.
“Scope was founded on the belief that we have a safer, faster, and more accurate way to measure the integrity of products and processes and can get that data to the right people in real time,” McCoy pointed out. “The result is that suppliers and end-users can respond before a failure incident occurs, as opposed to investigating what occurred after the fact.”
He added, “Industries also shouldn’t have to accept the level of risk that comes with manual visual inspection. With Scope, anyone can perform an expert inspection in minutes and make data-driven decisions about the safe use of the product with a level of consistency that is only achievable through automation.”
McCoy further explains, “(A standard for rope quality is) what we’re trying to enable. Our ground truth is the break strength, so we’ve broken thousands and thousands and thousands of rope segments in order to train our algorithms to understand that this rope and its visual features correspond to this break strength.”
Utility Stringing
As it turns out, there is an application where rope failure has dramatic consequences not unlike those seen in heavy lifting operations, and requires inspecting miles of rope for each operation.
“When they’re building new power lines, or replacing existing ones,” explained McCoy, “companies pull those power lines into place with extremely long lengths of synthetic rope. So, imagine you’ve got all of these empty towers—they literally take the rope and fly it out with a helicopter and attach it to each tower, and then tie it to the power line. They then pull the power line into place, which can be something like a four-mile-long segment of rope. That’s called utility stringing—they’re essentially stringing the power lines into place.”
McCoy and his colleagues eventually became aware of the fact that there are a lot of accidents connected to utility stringing. “Out in the field, the whole operation leads up to this moment when they pull those lines into place,” he indicated. “And if those ropes break, it’s like four-mile-long rubber band snapping, and everything gets destroyed. People even get hurt. So a lot of utility companies mandate that these ropes require inspection prior to the job—prior to the pull. But literally, up until now, it’s been a manual inspection. There will be four reels of rope on these large utility stringing machines, and it’s someone’s job every day, all day, to stand between two machines as they reel rope from one to another and visually inspect four miles of rope.”
McCoy explained that the rope is pulling through the stringing machines at maybe two miles an hour, because at any faster speed, the human eye can’t see it well enough to inspect it. And bear in mind, that inspector is only observing one side of the rope.
“So the error rate is really, really high,” he noted, “especially when they’re attempting to predict strength. I mean, humans are just not good at looking at a piece of rope and telling you how strong it is.”
In response, McCoy and his team released Control V1 in January of 2023. “It basically has three cameras inside of it—one at noon, one at four o’clock, and one at eight o’clock—(and) every five inches the images are taken as the line moves through [at up to four miles per hour] and then run across two different deep-learning neural networks. One predicts the break strength of that line segment based on a degree of abrasion, and the other identifies cut strands, splices, and debris. And then it assigns those damage points to the individual locations and builds a report automatically—all while the line is going through.”
Quality in Demand
According to McCoy, in addition to the obvious enhancement of the overall inspection process, and the resulting upgrade in safety and speed, the Scope system is extremely beneficial in helping users to identify areas in need of repair.
“Instead of just sitting there trying to remember, as it goes with a manual inspection,” he explained, “our system allows you to know exactly where you need to make the repair. After you do a scan, you see what line segments need to be addressed. And as it applies to the larger industry, perhaps the biggest value within the Scope system centers around how regulatory bodies try to mandate quality. They simply can’t do it. The only thing they can mandate currently, due to the standard practice of visual inspection, is that an inspection was performed.
“But they can’t mandate the quality of that rope because there isn’t a metric for quality. The only metric for strength is, really, what is the break strength? But you can’t find that unless you break the rope. Our technology has the ability to predict within five percent the actual break strength. So it enables companies to realize and maintain a safe level of compliance to the extent that when these utility companies come out and say you have to have at least a seventy-five percent break strength on your ropes, companies can actually fulfill that.”
McCoy also noted that Scope is looking at areas to expand. “It’s kind of the same issue with wire rope. They do have the MRT (Magnetic Rope Test), and it’s a service that typically requires a specialist. One of the best things about the tech we’re using is that we’re AI and Deep Learning first. As a technology stack, it can be applied to virtually any form of data—to pixels, point clouds, to text, to the feedback from an MRT device. So, we see a lot of areas where we could expand. Especially in the wire rope space.”
Additionally, the same technology incorporated into Control V1 can be incorporated directly into the application through an intelligent fairlead, which Scope has already developed. That tool embeds a stream of continuous inspection into any process every time a rope is paid out or reeled in, such as a crane.
Currently, McCoy and his colleagues have 13 active devices out in the field with customers, and another three on trial. “For now, we’re really targeting companies that maintain large fleets of this stringing line equipment because they have a high volume of inspection need,” he indicated. “They sometimes have to inspect before every operation, so we’ve got them planted at these facilities.”
Scope has inspected over twelve million feet of rope since January. At the end of the day, however, McCoy is content, for now, with keeping Scope’s game-changing technology in perspective. “We’re not trying to get people to demand our technology. We’re just trying to get them to demand a certain level of rope quality.”
For more information, visit www.visionbyscope.com. Additionally, you can download a complete case study at this link: White-Paper: Wagner-Smith Equipment Co. Implements AI Platform for Stringing Line Inspection
