Imagine starting your day knowing with 99% accuracy what it will entail. The exact time that you should leave your house, the best route to take to work (adjusted to changing traffic and weather conditions in real-time), and the driving speed to optimise your car’s fuel consumption. Your time of arrival at the office – even the number of people in the office lift, and whether you will need to take the stairs to be on time for that first meeting. With this level of predictability, you would be able to know how every decision you make affects your day’s goal (whatever that may be).
Similarly, the ability to predict an outcome in shipping – an industry challenged to decarbonise, reduce costs and embrace data transparency, can open the door to completely new possibilities. Shipping has always sought to optimise costs – but the drive to realise substantial savings has been hampered by a number of systemic barriers (split incentives between shipowner, operator and charterer or cargo owner being a key example). However, with regulatory and commercial pressure building day-by-day, optimising bunker consumption costs and minimising CO2 emissions are now essential for all stakeholders.
The rapidly growing volume of data available to the industry is increasingly allowing shipping organisations to seek technological solutions to difficult problems, improve efficiency, and uncover new opportunities. For those seeking to understand the fuel and carbon impact of any voyage, dynamic factors such as weather conditions, speed, route and currents combine with the particular condition of each vessel to create a complex puzzle. In the volatile environment of the sea, a very large number of factors affect a vessel’s operating performance at any given time.
“Predictive analytics” as a concept indicates the use of data, statistical algorithms, artificial intelligence and other tools to identify the plausibility of future outcomes based on what we know about the past. The objective is to go beyond knowing what has happened to providing a best assessment of what will happen in the near future. In shipping, this could mean predicting with a high level of accuracy how a voyage will go before the vessel has even left port, or proposing the correct actions (e.g. slowing down) en-route when the unexpected does happen (e.g. encountering higher adverse currents).
Huge volumes of bunker fuel are wasted every day due to inefficient sailing routes, speed policies and sub-optimal operation of vessel systems. Whilst minimising this loss is a continual mission of all companies, the industry has never had the correct tools to do so effectively. Take the traditional concept of “weather routing” as an example. Providers of weather routing solutions do exactly as their name implies: they create a route for a ship with the weather in mind. But they largely fail to account for one critical element – the vessel itself.
Knowing a vessel’s DNA adds vital layers of accuracy – and therefore savings – to the process of voyage planning. There is a profound difference between the capacity of traditional “weather routing” to produce a fuel-optimised route (based on pure weather data) and a next-generation approach that takes into consideration each vessel’s individual performance profile within the weather. At DeepSea, we call this revolutionary approach Performance Routing.
Here is one very tangible example of what this really means: a vessel with a clean hull will perform very differently to a vessel with a fouled hull despite sailing through the same weather conditions. Similarly, smaller vessels will be more affected by waves, and respond in a different way than larger vessels. At DeepSea we take a total of 19 parameters into account when we model any vessel.
Compared with traditional weather routing tools, an AI-powered performance routing platform such as DeepSea’s Pythia considers this sort of information vital in any voyage plan. As a matter of fact, Pythia is the world’s first voyage optimisation platform tailored to the exact performance of each vessel, under all conditions. Leveraging powerful AI models, Pythia learns to understand exactly how every individual vessel performs under the changing weather conditions and fouling state.
The Pythia platform, which is currently used on more than 300 deep ocean vessels of all types, uses weather data and vessel DNA to predict a ship’s performance profile with 99% accuracy – which results in an 8% to 10% reduction in fuel consumption.
Request a demo of Pythia here.
Explore the DeepSea platform live with one of our expert team members