Using AI to assess rope health 

Artificial intelligence (AI) tools have been created to simplify estimating the health of rope and to improve the speed and reliability of inspections. This helps to reduce accidents and injuries and enables owners to extend a rope’s operational life.

According to Samson Rope Technologies application engineering supervisor Michael Botterbusch, AI makes it easier to get repeatable, objective and actionable information about rope health, even when an in-person expert is not available.  

“To optimise rope service life and reduce operating costs and risk, it is critical to understand the current health and lifetime expectations of tug ropes,” he says. “The use of smart devices and AI guidance have greatly improved the speed and reliability of inspections.”  

Samson Rope uses AI tools to identify the status of tug ropes, compare inspection results and produce accurate rope health information.  

AI accuracy depends on the information inputs including rope and construction-specific data; its fibre and coating; and operations and time it is used, which all influence line performance lifetimes.

“By leveraging these advanced technologies and methodologies, organisations can extend the lifespan of tug ropes and achieve safer and more cost-efficient operations,” says Mr Botterbusch.  

“Ultimately, embracing AI in the inspection process serves as a pivotal strategy for maintaining rope integrity and operational excellence in maritime environments.”

Samson has tested used rope to assess its health for tug owners for over 25 years, so it has extensive historical data. Used ropes are inspected, pulled to destruction, and the remaining strength is compared to as-new information to generate a residual strength percentage.  

Comparative analysis

Samson assessed this database for trends among products, applications, and customers. Its analysis illustrated inaccuracies when relying simply upon rope usage numbers, such as hours in use or jobs conducted, to predict rope health without regular rope condition inspections.

For example, a rope may exhibit low residual strength after limited usage due to harsh weather or poor hardware maintenance. “Conversely, a rope may retain high residual strength after prolonged usage if subjected to particularly benign usage conditions.”

Samson found rope inspection results correlated well with residual strength. “Experience has shown that rope-condition assessment through inspections can be subjective in certain circumstances or subject to the experience of the rope assessor,” says Mr Botterbusch. But rope inspections are necessary to understand a rope’s remaining strength and life.

While rope condition is being determined, rope usage data can be tracked to understand rope condition and actions that might be taken to prolong its life.  

Samson found there is no correlation between number of towage jobs, or the period of time ropes are under strain, and its residual strength.  

This is because “all towing jobs are not the same,” says Mr Botterbusch, and “factors that damage a rope are more related to the details of each towing job instead of the total job count.” In the absence of supporting rope inspections, rope-usage history has a poor correlation to rope health.   

Inspections will focus on assessing abrasion on outer coatings; signs of external damage such as melting from frictional heating or partially cut strands; and looking at the inside surfaces of the hollow braid. There is a good correlation from plotting the external and internal abrasion ratings versus residual strength, says Mr Botterbusch.

Rising data volumes

“Combining the two forms of damage into a single equation and plotting the results versus the actual measured residual strength shows the power of this approach,” he says.

External abrasion indicates strength loss and a shorter field life for rope. A combination of rope inspection and testing can improve understanding of rope health, as would measuring tension during towage jobs

“Tension monitoring would be an excellent addition to a robust rope inspection programme but not a substitute for it,” says Mr Botterbusch. This data increases the volume of information to be assessed to provide an accurate view of a rope’s health, which AI can process more rapidly, efficiently and accurately.

Machine learning has been used to create classification models that can be used in concert with the inspection guide to rate external and internal abrasion levels. Samson is using AI to analyse historical images and data taken over the decades it has been testing used rope, together with the semi-quantitative inspection guide.  

“Using a proprietary image enhancement approach, objective ratings are produced that can be utilised by anyone with a mobile device and app,” says Mr Botterbusch. “However, the machine learning model is affected by the fibre type and rope construction and needs to be matched to the rope family being inspected.”