iHelm platform helps reduce variations in operations

Nowadays, vessel owners are curious and a bit confused about what information can be accessed from different equipment onboard. Also how they can be useful to reach their sustainability goals. The sky is the limit when it comes to data collection and information gathering. However, nobody has enough time to monitor various engine parameters or big data from vessels on fancy cloud dashboards 24/7.

This brings us to the question of how we can develop an intelligent product that learns from marine operations and delivers the best recommendations and insights to support decision-making. In this sense, Cetasol has invested heavily in artificial intelligence and data science technologies. This is used to bring the most valuable recommendations and insights to the fingertips of captains, managers, and vessel owners in real-time. Cetasol supports them in reducing fuel consumption & emissions and improving remote management.

Different segments

Vessels from different segments such as offshore operations, towage, or pilotage have different needs. Their configurations are also different with hardware including conventional marine diesel engines, electric engines, hybrid systems, complex navigation systems, generators, weather station devices, motion reference units, and so on. All these factors make the problem quite complex.

The flexible iHelm platform

iHelm system is flexible and can connect to different equipment via CAN-bus J1939, NMEA2000, NMEA 0183, Modbus TCP protocols, and analog signals. Accessing big data from different sources, analyzing them in a meaningful way, building the AI algorithms for providing the right recommendations to captains and vessel owners, and visualizing them in a user-friendly interface are the pillars of an intelligent product. Cetasol has a step-by-step approach to working with big data in different layers to avoid complexity.

Dynamic data models provided by iHelm

iHelm system creates separate dynamic data models for different data sources. They include the energy model, performance model, maintenance model, and operations model of the vessel. Since the system is compatible with any type of driveline, energy models can be generated for conventional marine diesel engines, electric motors, or hybrid systems. These models are trained with new data and updated after every operation. By doing so, we develop a system that learns from the operations and provides better recommendations and insights over time.

While real-time recommendations on a captain display help the captain improve his driving style and reach the destination on time with minimum fuel consumption. The best insights from operations can be useful for the management team in facilitating remote management and making better decisions. The AI models are being created frequently in the cloud and downloaded on the edge computing onboard unit. As the data becomes more consistent over time, the system provides better recommendations and insights.

digital twin of vessel. by cetasol.

Various AI models are being trained and updated with new vessel and operational data after each operation.

Our step-by-step approach to creating AI models can allow our customers to focus on their priorities, and make it easier to install & configure the system. Users can also expand the use of the system with additional data layers in time. These layers are grouped as:

iHelm system analyses vessel data in 4 different layers

As an example, you can consider a ferry that operates a similar route every day. Our experience shows that different captains operate in slightly different routes between the same two points. This situation leads to differences in fuel consumption up to 25%. On top of that, if you add the differences in speed and RPMs, fuel consumption between two captains on the same vessel can differ a lot. At Cetasol, we support to minimize this human factor which leads to high fuel consumption and emissions in marine operations.

A learning system

iHelm system collects engine data every day and machine learning models are being trained and updated to build a dynamic energy model of the vessel. During the operations, the system makes a comparison of the captain’s current fuel performance with the historical best and worst performances from previous trips that were completed in the same geographical region. Based on these analyses, the captain receives real-time recommendations with optimal speed and fuel consumption to reach the destination on time most economically. One of our customers, Öckerö in Sweden, has achieved 17% fuel savings with these recommendations in the first year of their subscription.

Follow us on LinkedIn to get our recent updates or receive our newsletter

Contact us on the form below if you want to know more about how we can help you reduce these variations that cost unnecessary fuel.