Route optimisation based on AIS tracks in the Baltic Sea
In the relatively shallow waters of the Baltic Sea, ships will have a higher fuel consumption compared to when sailing in, for example, the open waters of the Atlantic. This is due to higher resistance when the water depth under the keel is lower. We present the results of taking stored AIS tracks and optimising their route geometry as well as their speed to minimise fuel consumption, i.e. to minimise the shallow water resistance experienced by each ship. The results show key changes in traffic patterns, specifically fairway shortcuts that could be taken if there are no safety concerns.
The route optimisation analysis was part of the FAMOS Odin project, to which SSPA also contributed with an analysis of common Under Keel Clearance (UKC) in the Baltic Sea. Other tasks in the project, performed by fellow project partners, included bathymetric (water depth) measurements of large areas.
Access to a real-time feed of Automatic Identification System (AIS) data was provided for the project by the Swedish Maritime Administration. The data used in this analysis correspond to the whole of 2017. The dataset consisted of approximately 1.6 billion AIS messages, which were reduced to roughly 50% when filtered for commercial traffic only (AIS ship type categories 60 – 89).
The AIS data subset used in the analysis contained all tracks from tanker ships that travelled the area during 2017 (shown in the figure), although some tracks were excluded from the optimisation. This was either due to odd ship dimensions (too big or too small), which may have been wrongfully entered by the mariner onboard the ship, or the track crossing areas that were too shallow according to the bathymetric data, etc.
Internal SSPA software for route optimisation based on energy efficiency was used in this analysis. The software tool has been developed and used in both commercial and research projects. The software traverses a bathymetry raster and finds the most energy efficient route given a desired Estimated Time of Arrival (ETA). The analysis considers shallow water effects, based on ship dimensions and draught, as well as speed over ground. This study did not include Traffic Separation Schemes (TSS) nor effects from currents and wind, since the scope was to investigate routes in general.
The optimiser was set to require a UKC of at least 20%, i.e. a minimum UKC of 2 metres for a draft of 10 metres. Out of 19,159 AIS tanker tracks, 8,273 (approximately 43%) benefited from optimisation in fuel efficiency. Approximate fuel savings per track amounted to an average of 11% compared to the original track.
Although there was a theoretical 11% gain from optimising the routes, it is not possible for individual ships or operators to achieve such a gain. This is because there are several TSS areas in the Baltic Sea that ships must follow, and TSSs were not considered in the optimisation. However, from a regulatory perspective, shipping as a whole could benefit from moving – or adding alternatives to – the TSSs. As seen in the figure, from an energy efficiency point of view, the TSS east of Stockholm (area A) forces northbound traffic to sail further out east than necessary. From a safety point of view, however, there might be a higher risk in those waters, since there are numerous small islands and rocks in the vicinity.
The traffic south of Bornholm (area B), as seen in the figure, would benefit from taking a more northerly route according to the optimisation analysis, but still south of Bornholm if the route continues south of Denmark (Kadettrännan). If, however, the route continues into Öresund,
the route should remain north of Bornholm.
For south of the area between Gotland and Hoburg (area C), as can be seen in the figure, the greatest gain from performing optimisation was from not taking the route south of Hoburg when the route continues to the Gulf of Finland.
As stated earlier, if the optimiser is set to the same ETA, ships will arrive at the end of the optimised track at the same time as in the original track. These optimisation criteria have a significant effect on the results, considering that ships that travel on a shorter optimised route will reduce their speed, leading to double the fuel savings, resulting from both the shorter route length and the lower speed. If the optimiser is instead set to the same speed as in the original track, which can be understood to be a ship using a fixed service speed, then the results will be somewhat different: the optimiser will put more emphasis on finding deeper routes and less focus on the length of the route.
On a higher level, the difference in optimisation criteria relates to the contract specified for transporting the cargo. Ships either need to be at the destination at a pre-agreed time or will travel there at a fixed service speed. Optimising with a fixed ETA does the former and, with a fixed
speed, the latter.
The fact that the constraints of the optimisation, in this case with the fixed ETA, play such a significant role in the results also highlights the importance of understanding the implications that these constraints put on the result, since otherwise the optimised route could actually be a worse alternative than the original route. When weather and currents are included, as in the case of optimising planned routes in operational shipping, then understanding the implications of the optimisation constraints becomes even more important.
The illustration should be considered an example of what can be achieved with route optimisation. Ship operators can save fuel, but extra care must be taken to make sure that optimising is done under the right constraints and with accurate models and data. SSPA plans to continue to develop its existing route optimisation models and tools as part of future collaborative projects.
SSPA AIS database
SSPA has stored years of AIS data, from both Swedish coastal waters gathered by the Swedish Maritime Administration (SMA) and across Europe via the crowdsourcing initiative AIS Hub. At SSPA, we have the ability to work with all AIS data formats for customer-specific analyses.
Original tracks (red) and optimised tracks (green). Emphasis is placed on areas A, B and C. Illustration by SSPA.
FAMOS is an international project that comprises fifteen organisations from seven countries around the Baltic Sea. The project aims to map the water depth of previously unsurveyed areas used for commercial shipping in the Baltic Sea and to implement the BSHC HELCOM Re-survey Scheme. FAMOS also aims to help reduce emissions and the cost of fuel in the Baltic Sea by providing exact hydrographic data which can be used to re-evaluate the preferred routes used by vessels today, and to contribute to improved UKC awareness navigation by providing better data regarding water depth, the vessel’s vertical position and pre-calculated level of water surface at sea. www.famosproject.eu