Retrospective Analysis of Historical AIS Data for Navigational Safety through Recommended Routes

The HANSA project has investigated Retrospective Analysis of Historical AIS Data (Automatic Identification System) using both traditional AIS (terrestrial and satellite) and onboard AIS to enhance Navigational Safety through the concept of Recommended Routes. This has been done by collecting, combining and analysing the AIS data by AI supported algorithms. The aim was to develop a routing assistance tool to determine ship and environmental conditions and specific corridors based on the analysis of historic data such has AIS, hydrographic and meteorological information, considering national regulations. NAVTOR has in the first 18 months of the project (June 2018 ? June 2020) worked on the following elements; - Defining the concept of a ?recommended corridor? being useful in coastal/harbour areas as well in Global SOLAS trade. - Data collection and data deliverable consisting of historic AIS data and in spring 2019 also near real time AIS-feed, both sourced from onboard AIS-receiver on board NAVTOR connected vessels, and updated regularly (AIS time resolution is down to minutes, but in practical use now limited to 1-2 times a week) - Defining a prototype of the on board UI for Passage Planning and another prototype for the shore based web based UI (NavFleet), being NAVTOR?s new On Shore Web based platform for owners and operators. The first draft of the prototype is discussed with our customers, and user feedbacks are received. This will be an iterative process during the development of the prototype. - NAVTOR hosted the yearly HANSA project meeting in Bergen, August 27-28th 2019. - NAVTOR has in November received the first data sample (mesh) for recommended corridors. The mesh has been evaluated and commented from both the project and our perspective, and the feedback caused PUEB to refine the output. - During the spring 2020, NAVTOR has implemented results into our interfaces, and assessed it against our objective criteria for monitoring safe navigation. The aim has been to enable the mariner on board to detect potentially hazardous situations and operators ashore to monitor the general traffic situation as well to reduce risk in the planning phase. The conclusion of the intensive AIS analysis is that with AIS info alone one is not able to achieve sufficient quality in the «recommended corridors» when the goal is to be able to actively notify whether a ship is inside or outside such a corridor. Using the corridors to guide in and out of a port, such as Gydina in Poland which has been used in the Sprint project, can still be useful. A recommendation is to combine AIS info with additional info, e.g. include track from sailing ships. HANSA is funded through the MarTERA cofund partners German Federal Ministry of Economic Affairs and Energy (BMWi), Polish National Centre for Research and Development (NCBR) and Research Council of Norway (RCN).

The HANSA project has introduced the concept of "recommended routes". Project goals; -make it possible for the navigator on board to detect potentially dangerous situations when planning and monitoring. -provide safe routes for marine traffic based on sailing pattern of similar vessels. -will minimize the risk of e.g. groundings which can have calamitous impacts on the marine environment. -will enhance the safety by allowing the mariner to better predict the movement of other vessels in his surrounding area. -Shore side monitoring may utilize AI to predict dangerous crossing situations avoiding collisions and power groundings, saving the marine environment for e.g. oil spill. The Recommended route concept to be input to future projects dealing with autonomous shipping. The concept is unfortunately not yet ready to be used operational, but further work may finetune further; HANSA has laid the cornerstone for further utilization of AIS in combination of other information.

There are a number of data sources that can be used to support the generation of recommended corridors. The main challenge, which needs to be faced is that these data sets are large and complex. Traditional data processing tools are inadequate to process them within a tolerable time. As a response, Big Data technologies can be applied, which assume acquisition of vast amount of data from various sources and in different formats. This data is then processed, fused and analysed in (nearly) real-time. The project partners see significant market potential for additional modules or their product portfolio either for land-based VTS or CSS systems with the focus on operational use (e.g. anomaly detection) or onboard systems providing support in the route planning process, e.g. Navtor’s digital passage planning software NavStation. Each industry partner has a substantial customer base which can be exploited. As of today, no commercial system provides the smart combination of historical data (and thus navigational experience) with an operational system.