Predictive Maintenance – How Smart Vessels Take Care of Themselves

„We deeply regret that we have to cancel the departure due to unscheduled maintenance of the vessel.” This is clearly not what you want to tell passengers looking forward to a relaxing cruise trip or customers urgently awaiting goods to be shipped from overseas. Unplanned downtimes of modern cruise ships carrying almost 7,000 passengers or large container ships with a capacity of more than 20,000 TEU may not only lead to million-dollar losses but can also significantly damage a company’s reputation. Predictable availability and high reliability are crucial for success in a competitive sector facing volatile charter rates. Smart vessels can maximise their own availability while at the same time further reducing the risk of breakdowns.

Optimising Availability and Innovative Business Models

With traditional maintenance as it is still frequently used not only in the maritime industry, machines are overhauled or replaced at fixed intervals based on estimates and empirical values. Components are often replaced although they are still fully functional with the consequence of unnecessary docking times during which the vessel is not available. On the other hand, components that have already reached the end of their actual lifetime may not be replaced as they are not yet expected to fail based on their estimated lifetime.

Predictive maintenance uses an analysis of numerous data collected by sensors and monitoring systems to provide a reliable and precise forecast of when a machine or specific component actually needs to be maintained or replaced. Such forecast is inter alia based on temperature, usage, wear and tear, weather conditions and other factors in order to precisely determine when overhauls or replacements are required and to minimise maintenance time and cost.

A smart vessel will not only let the shipping company know as soon as there are any anomalies in its operational status. It may even book docking space, order relevant spare parts and provide the shipyard with all information required for a fast and efficient maintenance stop itself. New and innovative business models around predictive maintenance will surely emerge.

Challenges of Predictive Maintenance

If predictive maintenance offers such great opportunities, why is it that the majority of shipowners and shipyards have not implemented it already? Currently, the major hurdle seems to be data collection and its processing. It must be ensured that meaningful data of high quality is measured. Such measurement equipment requires considerable investment. For each vessel the influence of every component on the entire system must further be ascertained. A complex algorithm must be able to draw reliable conclusions from the operational status of a machine or specific component. A virtual model of the ship may help to visualise and conceptualise the data. Valuable experience may be derived from pioneering applications which have been analysing and using a multitude of data for years. No modern race car would work without the hundreds of sensors transmitting and analysing real time data commonly referred to as telemetry (which has been firstly implemented already in the 1980s and is used as derivate work product in more and more industries nowadays).

When data is collected and processed, data protection obviously becomes an issue. The data may contain sensitive information about the ship or even about crew members. Accordingly, restrictive regulations may need to be followed and agreements regulating flow of and access to the data must be put in place. Unauthorised access must be prevented.

Another question that needs to be addressed is liability. Who will have to bear the consequences if a predictive maintenance system fails? Is the shipyard, the supplier of the component or the company that developed the respective software liable for such failure? Without clear contractual risk allocation the shipowner may have to pay the bill itself at the end of the day.

However, all such challenges will be overcome, and predictive maintenance will become a more and more common measure in shipping to use resources in a more efficient way. Combined with artificial intelligence, it may soon be possible for a vessel not only to predict its failures, but also to fully look after itself and take care of its own maintenance. Certainly, one can expect that the ongoing digitalisation of maintenance will bring further change to the maritime industry.