Artificial Intelligence and Machine Learning offer value to the Shipping Community

By Mike Konstantinidis

 

The shipping industry is at a time of major change, having to deal with issues that are mainly relevant to the regulatory framework, while ship-owning and ship-management companies as well as charterers are required to make important decisions about their model of operation and strategic decisions on their business activity and investments. The optimization of environmental footprint of the cargo vessels becomes more and more pressing through ever increasing stricter regulatory obligations.

The development, however, which is revolutionizing the maritime sector has to do with the exploitation of all those data produced onboard the vessels and the extraction of the actionable intelligence that is hidden inside them. It seems that the management of the maritime data will be the field on which the new era of the industry will be built upon, giving significant competitive advantages to those who will perceive it promptly, prepare and adjust their operations accordingly.

Round the clock 24/7 data acquisition from all available sources provides a rich, high-resolution picture of ship performance rather than a patchwork of inconsistent, manual datasets and benefits add up to more than the sum of its individual technical, operational and cost gains, and depend on the user’s aims.

The volume of data produced by all kinds of equipment is so large at the time being that specialized techniques and methodologies have been implemented in order to handle, process and utilize them.  Machine Learning as the dominant technology that represents the concepts of Artificial Intelligence, consists of a set of algorithms, tools and techniques, optimized to handle big data in a way that mimics the human learning process. More specifically, the process involves the utilization of existing datasets to develop models that will learn to recognize trends and patterns that couldn’t be noticed with other methods, in order to be able to identify and predict future behaviors. Significant part of the process is that the models are retrained by themselves, thus shelf-improving their accuracy and intelligence.

However, the greatest innovation that offers an invaluable competitive advantage to the shipping companies has to do with the prediction of future behaviors and events. As an example, one can imagine how useful would be the accurate projection of fuel consumption of a specific vessel during the next voyage through a specific route, considering factors such as the required speed, the weather conditions in the whole path, the actual performance of the main engine, the current state of the hull, etc. Functionality like this can only be provided by the proper use of Machine Learning technology.  The correlation of shaft power generated by the main engine of a vessel and the fuel consumption in various vessel speeds is not described by any mathematical engineering formula.  Such results create a great tool that facilitates the work of shipping companies, increases productivity, and provides significant help in the effective control of operating costs as well as the monitoring of the regulatory framework.

In practice, the solutions powered by such digital technologies essentially set the artificial intelligence in the service of the shipping community. Any shipowner monitors in real time a wide range of vessel parameters, records thousands of signals, combines data from third party services (weather providers, AIS) and is able as using a “virtual personal assistant” to make valuable conclusions, reply to ad-hoc queries and requests of any kind, send notifications about critical events and prepare personalized per user/department reports. 

The overabundance of data which are available to the shipping industry, creates indeed significant challenges. Digitalization and Shipping sector should be willing to combine in the most comprehensive way, maritime know-how, naval/engineering expertise and modern digital technologies to provide to the crew and the shipping community a clearer perception of the environment in which they operate.

METIS Cybertechnology, a company specializing in the development of innovative solutions for the global Maritime Community, is one of the greatest ambassadors for the implementation of Machine Learning in the shipping industry. METIS' objective is to ensure an integrated and reliable process of data collection, real-time performance monitoring and intelligent analysis, providing useful and actionable information addressing the needs of various roles, departments, and levels of management in a shipping company.

As the industry’s only end-to-end platform powered by AI, however, the true value of the METIS solution is cumulative for individual ship efficiency and fleet-wide. Using machine learning, the platform evaluates itself every seven days and retrains monthly to refine the correlation between weather, hull fouling, power use, fuel efficiency and other significant parameters.