In the past, computers couldn’t learn from their experiences – they needed to be told exactly what to do. This made data analytics very basic, as computers had to be instructed to consider ‘X’ if ‘Y’ had happened.
GREENSTEAM IS MACHINE LEARNING
Deep insights with accurate, actionable advice and measurable financial gains.
WHAT IS MACHINE LEARNING?
Today we can get a computer to learn from experiences and that’s exactly what machine learning is all about – using past experiences to model the future.
01.We can train ‘the machine’ to use a wide range of data sources to make incredibly accurate predictions about what is going to happen next - even if it has no direct experience of the situation.
02.The larger the volume of data, parameters and interactions it has access to, the more it can learn, predict and offer advice related to future scenarios.
03.No human could process the amount of data a machine can, and certainly not with the same split-second speed, accuracy and reliability.
04.A machine never forgets; it never dismisses any information as insignificant; and it has no distractions. Machine learning is a type of artificial intelligence and it’s being used today in almost every industry.
WHY MACHINE LEARNING?
Using historical vessel data, our machine learning technology creates a performance model of the vessel. The greater the volume of data the performance model is fed, the more it can learn and the more accurately it can make predictions on how the vessel will perform in the future.
The performance model, created using vessel operational data and AIS, weather and sea-state data, is the basis for how GreenSteam identifies the inefficiencies, however small, that contribute to your fuel wastage.
Efficiency gains can only become a reality if all the complex operational variables from a vessel, and how they impact each other, can be identified, optimised and measured.
Conventional technologies cannot possibly understand the way they interact and what impact this has on vessel performance.
But machine learning can.
Machine learning systems improve their performance through direct experience of real-world conditions.
Through every vessel journey, irrespective of the weather, sea-state, speed, load or route, GreenSteam Discover is learning. It’s adapting and updating the performance baseline model so that it can deliver deeper insights into the vessel’s operational efficiency and provide ever more accurate, real-time and predictive advice on how to improve it.
Through every vessel journey GreenSteam Discover is learning, adapting and updating the performance baseline model, so that it can deliver deeper insights and more accurate advice.
HOW DO WE USE MACHINE LEARNING?
GreenSteam uses machine learning to systematically aggregate vast amounts of complex data from sources on and off vessels such as fuel quality and usage, hull coating and cleaning schedules, costs and outcomes, speed, weather, sea-state and route. This happens in an automated way and requires no human intervention. From this data, an operational performance model of the vessel is created and it is this model that delivers the efficiency insights from which actionable, predictive advice is made.
From advice on vessel and voyage trim, to speed and route. From insights into hull coatings performance to the efficacy and costs of cleaning and coating schedules, GreenSteam machine learning technology delivers highly-accurate, actionable insights into where large and small efficiency gains can be made, either as predictive advice on how to operate your vessels more efficiently, or as dynamic advice in real-time to the crew on the bridge.
BENEFITS BEYOND FUEL SAVING
The GreenSteam machine learning approach is not just about fuel saving – there are additional sources of value to be had from your vessel data across all your fleet operations. the nature and flexibility of the machine learning approach means we can incorporate a wide range of operational factors such as early/late arrival fees, staff overtime etc., into the analysis. Through these additional factors, we can model scenarios that provide insight into operational areas that other methodologies cannot reach.