Back to Newsroom

Signal's forecasting algorithm gains a USA patent

This article examines recent progress and initiatives by the shipping industry to meet new IMO targets and provides Signal Ocean Platform data on emissions developments and trends.

The Signal Group
April 2, 2024

Trying to form accurate predictions of commercial ships seems easy enough to the sound of it, they are pretty slow beasts anyway. But if the predictions need to be commercially valuable i.e. months ahead over multiple voyages, then it becomes a quite challenging and novel problem, with no ready made solutions. 

We addressed this challenging problem by turning it into two different sub-problems: first, define the search space, and second derive a reward function that gets its maximum value over the best solution in that space. Both problems turned out to be of similar difficulty. 

For the first sub-problem you can actually try the brutal force solution of any possible ship voyage, but estimating anything more than a single load-discharge pair becomes exponentially difficult i.e. a no-go solution. Even cutting it down to all possible solutions for a particular ship at a particular location leaves you with a prohibitively high number of possible solutions. To address that we utilized a Graph-Theoretic approach over which we find the most probable paths over all the data points (i.e. particular commercial info regarding the ship at hand). For example, if the only data point that we have is that the next discharge of a VLCC tanker is going to be in China we can estimate that its loading port is going to be somewhere in AG so we form a “load-AG discharge-China” prediction. Adding to that the output of AI models we can pin down this prediction to specific ports. Applying this data-algebra over all the data inputs we can generate a set of solutions that is guaranteed to contain the optimal one, albeit being of small size. 

Having generated the search space is only half the story. The story is not complete until you select the one solution in the set that best matches your data, taking into consideration that the data itself is “untrustworthy”, “conflicting”, “noisy”, “ambiguous” and “incomplete”. In other words, all the bad data words you can find in all of the data science papers combined. Again here using a lot of business knowledge and appropriate handling of all data formats, that range from gps readings of the ship’s locations to ship open ports advertised over “what’s app”, we designed a sorting algorithm that selected the best path more than 95% of the time. More than 40 different statistical models and advanced algorithms have been combined to reach such a high level of accuracy. 


All this has come together into our first USA awarded patent (link1, link2). To get a first idea of the commercial result,  sign up to the Free edition of the Signal Ocean Platform.

Creating a sustainable world requires us to embark on a journey towards a zero emission future, where every step is a commitment to preserve our planet for future generations.
Albert Greenway
Environmental Scientist, Sustainability Expert
No items found.

Increased Use of Renewable Energy:

Shipping companies are embracing renewable energy sources to power onboard systems and reduce emissions during port operations. Solar panels and wind turbines are being installed on vessels to generate clean energy, reducing reliance on auxiliary engines, and cutting down emissions. Shore power facilities in ports allow ships to connect to the electrical grid, eliminating the need for onboard generators while docked.

Collaboration and Industry Partnerships:

Recognizing that addressing emissions requires collective action, shipping companies, governments, and organizations have formed partnerships and collaborations. These initiatives focus on research and development, sharing best practices, and promoting knowledge transfer. Joint projects aim to develop and deploy innovative technologies, improve infrastructure, and create a supportive regulatory framework to accelerate the industry's transition towards a greener future. The Zero Emission Shipping - Mission Innovation.

To pave the way for a greener future in shipping, the availability of alternative fuels plays a vital role in their widespread adoption. However, this availability is influenced by factors such as port infrastructure, local regulations, and government policies. As the demand for cleaner fuels in shipping rises and environmental regulations become more stringent, efforts are underway to improve the accessibility of these fuels through infrastructure development, collaborations, and investments in production facilities.

Liquefied Natural Gas (LNG) infrastructure has seen significant growth in recent years, resulting in more LNG bunkering facilities and LNG-powered vessels. Nonetheless, the availability of LNG as a marine fuel can still vary depending on the region. To ensure consistent availability worldwide, there is a need for further development of LNG supply chains and infrastructure. For biofuels, their availability hinges on production capacity and the availability of feedstock. Although biofuels are being produced and utilized in various sectors, their availability as a marine fuel remains limited. Scaling up biofuel production and establishing robust supply chains are imperative to ensure wider availability within the shipping industry.Hydrogen, as a fuel for maritime applications, is still in the early stages of infrastructure development. While some hydrogen vessels have been tested or introduced in the first quarter of last year, the infrastructure required for hydrogen production and distribution needs further advancement.

Ammonia, as a marine fuel, currently faces limitations in availability. The production, storage, and handling infrastructure for ammonia need further development to support its widespread use in the shipping industry.Methanol, on the other hand, is already a commercially available fuel and has been used as a blend with conventional fuels in some ships. However, its availability as a standalone marine fuel can still be limited in certain regions. Bureau Veritas in October 2022 published a White Paper for the Alternative Fuels Outlook. This white paper provides a comprehensive overview of alternative fuels for the shipping industry, taking into account key factors such as technological maturity, availability, safety, emissions, and regulations.

Creating a sustainable world requires us to embark on a journey towards a zero emission future, where every step is a commitment to preserve our planet for future generations.
Albert Greenway
Environmental Scientist, Sustainability Expert

Increased Use of Renewable Energy:

Shipping companies are embracing renewable energy sources to power onboard systems and reduce emissions during port operations. Solar panels and wind turbines are being installed on vessels to generate clean energy, reducing reliance on auxiliary engines, and cutting down emissions. Shore power facilities in ports allow ships to connect to the electrical grid, eliminating the need for onboard generators while docked.

Collaboration and Industry Partnerships:

Recognizing that addressing emissions requires collective action, shipping companies, governments, and organizations have formed partnerships and collaborations. These initiatives focus on research and development, sharing best practices, and promoting knowledge transfer. Joint projects aim to develop and deploy innovative technologies, improve infrastructure, and create a supportive regulatory framework to accelerate the industry's transition towards a greener future. The Zero Emission Shipping - Mission Innovation.

To pave the way for a greener future in shipping, the availability of alternative fuels plays a vital role in their widespread adoption. However, this availability is influenced by factors such as port infrastructure, local regulations, and government policies. As the demand for cleaner fuels in shipping rises and environmental regulations become more stringent, efforts are underway to improve the accessibility of these fuels through infrastructure development, collaborations, and investments in production facilities.

Liquefied Natural Gas (LNG) infrastructure has seen significant growth in recent years, resulting in more LNG bunkering facilities and LNG-powered vessels. Nonetheless, the availability of LNG as a marine fuel can still vary depending on the region. To ensure consistent availability worldwide, there is a need for further development of LNG supply chains and infrastructure. For biofuels, their availability hinges on production capacity and the availability of feedstock. Although biofuels are being produced and utilized in various sectors, their availability as a marine fuel remains limited. Scaling up biofuel production and establishing robust supply chains are imperative to ensure wider availability within the shipping industry.Hydrogen, as a fuel for maritime applications, is still in the early stages of infrastructure development. While some hydrogen vessels have been tested or introduced in the first quarter of last year, the infrastructure required for hydrogen production and distribution needs further advancement.

Ammonia, as a marine fuel, currently faces limitations in availability. The production, storage, and handling infrastructure for ammonia need further development to support its widespread use in the shipping industry.Methanol, on the other hand, is already a commercially available fuel and has been used as a blend with conventional fuels in some ships. However, its availability as a standalone marine fuel can still be limited in certain regions. Bureau Veritas in October 2022 published a White Paper for the Alternative Fuels Outlook. This white paper provides a comprehensive overview of alternative fuels for the shipping industry, taking into account key factors such as technological maturity, availability, safety, emissions, and regulations.

Ready to get started and outrun your competition?

Request a Demo
BACK TO STORIES

Signal's forecasting algorithm gains a USA patent

Posted by
The Signal Group
|
April 22, 2021

Trying to form accurate predictions of commercial ships seems easy enough to the sound of it, they are pretty slow beasts anyway. But if the predictions need to be commercially valuable i.e. months ahead over multiple voyages, then it becomes a quite challenging and novel problem, with no ready made solutions. 

We addressed this challenging problem by turning it into two different sub-problems: first, define the search space, and second derive a reward function that gets its maximum value over the best solution in that space. Both problems turned out to be of similar difficulty. 

For the first sub-problem you can actually try the brutal force solution of any possible ship voyage, but estimating anything more than a single load-discharge pair becomes exponentially difficult i.e. a no-go solution. Even cutting it down to all possible solutions for a particular ship at a particular location leaves you with a prohibitively high number of possible solutions. To address that we utilized a Graph-Theoretic approach over which we find the most probable paths over all the data points (i.e. particular commercial info regarding the ship at hand). For example, if the only data point that we have is that the next discharge of a VLCC tanker is going to be in China we can estimate that its loading port is going to be somewhere in AG so we form a “load-AG discharge-China” prediction. Adding to that the output of AI models we can pin down this prediction to specific ports. Applying this data-algebra over all the data inputs we can generate a set of solutions that is guaranteed to contain the optimal one, albeit being of small size. 

Having generated the search space is only half the story. The story is not complete until you select the one solution in the set that best matches your data, taking into consideration that the data itself is “untrustworthy”, “conflicting”, “noisy”, “ambiguous” and “incomplete”. In other words, all the bad data words you can find in all of the data science papers combined. Again here using a lot of business knowledge and appropriate handling of all data formats, that range from gps readings of the ship’s locations to ship open ports advertised over “what’s app”, we designed a sorting algorithm that selected the best path more than 95% of the time. More than 40 different statistical models and advanced algorithms have been combined to reach such a high level of accuracy. 


All this has come together into our first USA awarded patent (link1, link2). To get a first idea of the commercial result,  sign up to the Free edition of the Signal Ocean Platform.

Latest Stories

In the news

Latest Press Releases