During the 2019 SMART4SEA Conference, Mike Konstantinidis, CEO, METIS Cyberspace Technology, described how Artificial Intelligence and Machine Learning can apply in shipping and improve operations. As such, he talked about the METIS Virtual Agents, the first cyber personal assistant for vessel monitoring & management, interacting with its colleagues in the maritime company via natural language processing, calculating KPIs, and making complex prognostics and diagnostics.
A shipping company managing a fleet is under pressure and relentless demand to sink the costs and reduce the fuel oil consumption, optimize the routes, and answering special questions triggering shipping company executives. This creates an extreme complexity and a huge correspondence and communication between each other and the crew. Additionally, one thing we need to consider is the apparent limitations of the crew capabilities.
Those questions, triggering the shipping executives, are varying from very simple, to extremely complicated ones that need a lot of work to address. In order to be able to answer those questions, first you need to have data, either from the vessels even from fragmentary information that currently the shipping industry keeps, and third-party data like weather. You need to ensure that this data has the quality to be utilized. You need to properly format them, in order to be able to make some proper decision making, and of course to have the necessary expertise and knowledge on subject.
In METIS, we have created the first cyber personal assistant monitoring the fleet and staying at the disposal of every employee of the shipping company. She can work 24/7, is able to monitor the fleet 24/7 since this is a non-stop business activity. It remains at our disposal to answer any question, spontaneously can send alerts or reports either timely or on demand, can perform analysis on specific subjects, either technical or operational.
Let’s see an overview of the whole METIS concept. First of all let’s concentrate on how we do the data acquisition from the vessels. We designed, developed and produced, intelligent devices based on our unique state-of-the-art technology that work as collectors of the signals from any sensors on board. This is mostly essential since shipping industry lacks standardization and there are so many different proprietary protocols and interfaces that need to be supported. Our devices and the whole concept gathered type approved by the Lloyd’s Register classification society, for collection, transmission and analysis of the vessel performance data. Those devices are both wired (Ethernet) and wireless so the installation is very fast. In two working days, we complete a full installation on board the vessels, and they have the state-of-the-art, patent pending, wireless technology as well, so they don’t need any cables. We collect data from any data source on board the vessel, either navigational or main engine or steam production, it depends on the type of the vessel. But this is simply not enough. We need to enrichen this data feed with third party data (weather, AIS, etc) and even from manually taken data, like the noon reports or other information that is at our disposal.
Then, we undertake the duty to transform those data into actionable information for the users. There are two aspects of this interaction. The first one is what we would call quantitative analysis. This is mainly serving the technical departments, the technical staff of the shipping companies would like to see some kind of data visualization, after those data are processed, cleaned, filtered, in order to be able to be leveraged properly. We use industry-standard applications like Tableau, or Microsoft Power BI or others, in order to be able to provide the information in a utilizable way, compared to the reference lines, weather and ISO corrected.
Here comes the Artificial Intelligence to be involved. This is how Artificial Intelligence augments the data acquisition. For example, say that we have a torque meter connected and somehow this torque meter goes wrong or is uncalibrated. We get the turbocharger RPMs, the main engine RPMs, the scavenge air pressure and fuel pump mark, and our algorithm can calculate a quite accurate estimation of the shaft power. Another example would be how we approach the reference lines of various equipment, machinery installed on board the vessels. After some time that we get data, it is very easy for the technology to understand and identify when a measurement is somehow wrong or unjustifiable.
Then goes to the qualitative information. Here, in order to bypass the difficulties and the complexities that shipping industry is suffering, we created the worldwide first chatbot for shipping. We created a user functionality that the humans interact with the system in plain English. You can just ask questions and immediately get answers in any of the internal collaboration platforms that the shipping company adopted and uses every day. It can be your e-mail, or Skype, or TEAMS, or others. So we have METIS as a partner, a colleague of our company monitoring and digesting thousands of measurements. Every 15 seconds she gets an indication, a measurement from every data source, and all these enrichened by third party data and processed properly. You have a personal assistant that works for you 24/7, can answer any questions or send you an alert in your e-mail, or a report that is fully completed and comprehensive at your wish by the intelligence created by our technology.
On the other side, Machine Learning technologies are already creating mature and reliable results, as far as the prediction is concerned. After some months, maybe 4 to 8 months of getting rich and versatile data, our technology can predict with considerable accuracy, how some specific parameters, will behave in the future. This is the way how technology will augment us, increase our efficiency and productivity, and guide us and help us to go to the next day of the digital transformation of the shipping sector.
Above text is an edited version of Mr. Mike Konstantinidis presentation during the 2019 SMART4SEA Conference.
View his presentation here.
The views presented hereabove are only those of the author and not necessarily those of SAFETY4SEA and are for information sharing and discussion purposes only.
About Mike Konstantinidis, CEO, METIS Cyberspace Technology SA
Mechanical Engineer (BSc, MSc), having gained a broad range of skills for over the years, heavily focused to the Maritime Sector. Holding over 25 years of insightful experience in Corporate Governance, Project Management, Sales & Marketing, Mergers and Acquisitions, Change Management and International Business Development with wide range of business and industry verticals. Effective communicator, with ability to accomplish marketing plans improving brand awareness, designed and executed operational plans and business strategy creating revenue and profitability growth, with emphasis on sustainability.
