There’s a reason digitalization is the industrial buzzword of our time. Artificial intelligence and big data analysis are fundamentally changing industrial processes and generating big headlines along the way. AI is now helping researchers develop new medicine, to choose one recent high-profile example. In the marine industry, autonomous vessels are closer than ever, and already they’re generating plenty of headlines.
But real transformative breakthroughs do not always occupy a large place in the public eye. Track and trace and real-time analytics have quietly added a level of reliability to supply chains that was unimaginable a generation ago, changing how packages are shipped and even how food gets to the supermarket. Similarly, in the marine industry, condition monitoring has changed maintenance from reactive to proactive, lowering maintenance costs and increasing safety.
A key part of a crewmember’s job, observing how machines function, is now often assisted by an increasing number of sensors. The shift has been slow sailing for the marine industry, but the cost of sensors and data transmission has dropped to a level where smart condition monitoring is now widespread. Sensors that measure vibrations or temperature produce concrete data at levels of precision that humans cannot.
But the ability to take the observations these sensors capture and turn them into actionable information is the big breakthrough of the data movement. This breakthrough has come in two steps.
The first has allowed for big data analysis to diagnose issues. More recently, condition monitoring solutions have been able to take things a step further, offering AI-enabled prognostic help as well.
Sirius, a Swedish shipping company, is a good example of condition monitoring helping with diagnostics. Sirius’s vessels are equipped with sensors that are connected to SKF’s Remote Diagnostics Centre in Hamburg. There, computers can analyze vibration data in real time. Should they notice an anomaly, Sirius maintenance personnel are alerted and investigate further. In this way, Sirius is able to catch potential issues before they cause unexpected stoppages and breakdowns. “Over the lifetime of the equipment we expect to save a lot of money,” Benjamin Fhager, the company’s Technical Coordinator, said in 2018.
Larger shipping companies have brought this technology in-house. In 2018, Carnival opened a new Fleet Operations Center in Miami that performs real-time condition monitoring for its fleet of nearly 30 cruise ships, among other things, like tracking and logistics support.
In other industries, prognostic, or predictive maintenance, has taken big data analytics to another level. The marine industry is following suit.
Machine learning and deep learning
Presenso is an SKF company that uses machine learning and deep learning, two kinds of AI, to do more than interpret sensor data and diagnose problems in real time. It also automatically compares the data to historical information on performance and wear and tear to predict the continued lifespan of a given piece of equipment.
As more data is fed into AI solutions like Presenso’s, they learn to become even more accurate with their predictions. The predictive abilities offer users the foresight to prepare for maintenance requirements ahead of time, ensuring spare parts are available when needed, reducing downtime and the need to stockpile expensive parts, preventing potential failures, and increasing safety.
“Changes in vibrations can give us a lot of information about performance,” said SKF’s Ruth Eickhoff, vibration specialist in the Global Remote Diagnostics Centre in Hamburg. “The more data these tools have, the more they learn, and the better their predictive capabilities become.”
Diagnostics and even prognostic analysis do not have to be big-budget operations. Crews aboard Seanergy Maritime Holdings Corp’s fleet use a handheld sensor developed by SKF to collect vibration data from machinery, like pumps, purifiers, electric motors, fans, compressors, shaft generators, and blowers. The data is analyzed off-ship and compiled into regular reports that the company uses to inform maintenance decisions and scheduling. Though it lacks the real-time element of connected solutions, the reports nevertheless allow for a level of foresight that shortens maintenance windows, saves money, and increases safety.
Seanergy’s Technical Manager suggested such predictive capabilities would be the beginning of a cultural shift in the industry, moving away from a “firefighting” mentality and more toward prevention.
That really is smart. Isn’t it?