02 May The new era of weather routing
Fundamental technological evolutions indicate that the weather routing market is ripe for disruption.
Written by Anouar Kiassi, CEO.
From weather routing to voyage optimisation
For years, the weather routing field was dominated by decades-old weather companies. These companies have developed (or acquired) software solutions they put on top of their weather models and consulting activities to provide routing services.
First, let’s clarify the meaning of weather routing. Initially, weather routing was about helping ships avoid severe weather conditions (e.g. high waves, storms, etc.). Today, in addition to safety, which remains the most critical factor, weather routing takes into account multiple operational, economic and environmental criteria to optimise the voyage. Therefore, ‘voyage optimisation’ seems to be the most appropriate name for this service. However, ‘weather routing’ appellation is still widely used in the industry to refer to ‘voyage optimisation’.
It goes without saying that the best voyage optimisation solution must combine a good mathematical algorithm and a good weather forecasting model, as well as a good ship-weather response model. On one hand, a very smart algorithm will give wrong predictions and advice if run on top of low quality weather forecasts. On the other hand, an unreliable algorithm will provide poor recommendations whatever the quality of weather forecasts is.
The technical strategies to build weather routing software and services have evolved over the years. The outcome reflects the state of the technology at the time of the design, but also the strategic choices made by each company.
In this article, we will compare the traditional weather routing companies and the new generation voyage optimisation companies, and look at the technological developments that will disrupt this market and reshuffle the historical positions.
Weather forecast
The traditional weather routing companies have, almost all, started as weather companies. They spent years developing their own weather models and put in a lot of effort to maintain them. While this can be seen as an intrinsic strength, it can also be a serious constraint to innovation.
At Ascenz Marorka, we consider raw weather data as a commodity. In fact, if many companies can develop a proprietary weather model, most of the raw data used to feed the models is coming from public or semi-public organisations (e.g. ECMWF, satellite imagery) because of the high investment needed to acquire this data. When the data is coming from private companies, it is made available for everyone with the same pricing model. Therefore, there is no decisive competitive edge there for weather companies.
Besides, the experience and the scientific literature show there is not a single model capable of describing accurately all oceanic and atmospheric conditions globally. It is widely recognised that some models are strong in some areas and weak in others. More importantly, some areas are so specific that they require high-resolution models (in time and space) to describe the weather conditions.
Furthermore, new modelling paradigms based on data and AI are being developed with promising results and high disruptive power (for example, the Google Deepmind algorithm was able to provide faster and more accurate weather forecasting).
At Ascenz Marorka, we believe that all models can be improved and all paradigms can be disrupted. Therefore, we created an innovative process that combines the best-in-class models and data sources. We have proven that the result is better than any model taken separately.
Routing algorithm
Voyage optimisation for ships is not a theoretical mathematical problem. It requires an extensive and continuous dialogue with experienced mariners to make sure the recommendations are relevant. In fact, the requirement that is expressed the most often by our customers is ‘trust’. In order to build that trust, we try to make our recommended routes ECDIS-ready wherever possible. This requires the integration of many factors in the algorithm: the experience of mariners, navigational databases, operational constraints, ship design constraints, etc. This way the routing solution can act like an assistant to the shipmaster.
Ship-Weather response model
Vessel-Weather response modelling, especially in rough seas, is a critical component to demonstrate reliability of the system and gain the users’ trust, because, in reality, ‘calm sea’ is a rare condition.
A poor modelling of Vessel-Weather responses has many operational consequences. On one hand, over-estimated speed-losses generate large deviations and therefore overconsumption of fuel. On the other hand, under-estimated speed-losses lead to unsafe direct routes through severe weather coditions and therefore to last minute deviations, which cause additional costs or delays. With our experience, we came to the conclusion that only a strong combination of data-driven and advanced physical modelling can bring a well-balanced and realistic solution.
New generation weather routing
There are many aspects that differentiate modern route optimisation solutions vs. traditional solutions.
Automation
It is not news to anyone that crews are overwhelmed with administrative tasks. With regulations piling up, this situation will not get any better, any time soon. Therefore, automation through digitalisation is the quickest answer to that challenge.
Many traditional weather routing providers use extensive teams in low cost countries to provide passage plans to masters. While this looks like a tailor-made service, it actually requires many iterations with the master to reach an acceptable route. In a traditional sequential process, the provider suggests a first route. Then, many iterations with the master follow to take into consideration the safety requirements. At the end of the process, the route deviates from the optimum in terms of energy/emissions savings.
Modern weather routing solutions integrate safety constraints in the performance optimisation engine. Therefore, the optimisation process takes into account both conditions at the same time and not one after the other, leading to a safe and optimised ECDIS-ready route. Furthermore, route updates and weather bulletins are fully automated and provided more frequently. That level of reactivity cannot be matched with a human service at a reasonable cost. Therefore, the role of the service team changes. It focuses on handling rare exceptions, complex route and dangerous weather situations. The highly experienced and qualified service team brings real expertise to the users (weather, navigation, safety, performance, alternative fuels, alternative propulsion systems, etc.). They can also deliver consultant class services to the shipping companies at a higher level in order to improve their internal processes and procedures, leading to a wider and more effective deployment of the best practices across their fleet and operational departments.
Cloud-first
Traditional weather routing solutions have all started with on-board programs because the connectivity on-board was a challenge at the time. While this option mitigates the risk of poor or unstable internet connections, it comes with many shortcomings:
- Installation or troubleshooting may require the attendance of a technician on-board
- Installation requires the involvement of the shipping company’s IT department
- Updates require sending a physical media on-board (e.g. CD-ROM). In reality, because of the practical challenges this operation poses, updates are not performed very frequently and the crew uses outdated versions.
- Low resolution and updates for weather forecasts to accommodate the limited connectivity
- Low resolution geographical grids to cope with the limited computational power on-board
- Computer specific bugs, as the software is running on all types of computers.
Modern weather routing providers are forward looking and base their solutions on cloud technologies. In fact, with the fast development of the hyperconnectivity on-board (e.g. Starlink), it is just a matter of time before most ships have a high-speed internet connection and become just like an office onshore.
Cloud-native solutions offer significant advantages:
- Immediate deployment with no need for any attendance or installation on-board. The software requires a standard web browser that exists by default on any computer
- Frequent and instant updates for software components, the maps, the navigational databases and the weather forecasts
- Unlimited computational power allowing the integration of high-resolution data sources
- Most of the time, the weather routing solution is seamlessly integrated with other modules, offering advanced vessel performance monitoring and optimisation.
Many traditional weather routing solutions created cloud extensions. However, as they could not spend the same effort on two fronts, the online versions usually provide very limited capabilities if not only a repository of manually created passage plans or performance reports.
Dynamic vessel performance modelling
Some traditional solutions route all vessels alike. They do not take each vessels’ specifics into account for better performance simulation and optimisation. Comprehensive vessel modelling provides valuable information regarding the vessel behavior in terms of fuel consumption (speed-consumption tables), reaction to weather factors (over/under consumption and ship motion risk assessment), multi-fuel management, energy-saving device management, alternative propulsion systems usage, and impact of hull and propeller fouling/maintenance.
That said, the shipping companies do not always have the data to allow a comprehensive modelling of the ship. Therefore, a progressive approach is required. This goes from physical modelling, to noon-reports-based modelling, to sensor-data-based modelling. Dynamic modelling is the process of starting the model building with whatever data is available and improving it over time, as we collect more data from existing or new sources.
API and interoperability
Weather routing is usually delivered though a software or a service team. Very few routing companies provide APIs. This allows customers to integrate weather and weather routing capabilities in their existing software, making them part of their streamlined operations. Depending on the use cases, the API can provide fast and statistical simulations for long term planning tasks, or very detailed and accurate simulations for short-term operational tasks.
Ascenz Marorka’s weather routing technology
Ascenz Marorka’s technology was born from a detailed assessment of market solutions. Therefore, we have created a modern, state-of-the-art, forward-looking solution that incorporates many innovations at all levels:
- Data, with high frequency data sources for weather and safety
- Weather modelling, with combined models
- Ship modelling, with dynamic modelling
- Multi-objective and multi-criteria algorithms answering all use cases
- Constant and variable power algorithms for both thermal and wind propulsions.
- User experience designed by mariners, for mariners
- Highly qualified service teams focusing on expertise, not low value repetitive tasks
- Automation and cloud native
- API ready.
Want to request a demo? Let’s get in touch: Contact Us – Ascenz Marorka
