International Transportation | Collection 2022 25 CADMUSS - an innovative project to improve maritime safety Collision prevention, Safety, Maritime transport chains, Ship encounters The evaluation of a (maritime) traffic situation requires sound training and professional experience. Decisions can be made based on this training and experience. (Partially) autonomous ships must be trained or require generalized algorithms to react appropriately in any situation. The goal is for vessels to be able to determine the technical manoeuvring distance and the required personal perceived safety distance. Sönke Reise, Carsten Hilgenfeld, Diego Piedra-Garcia A utonomous or semi-autonomous vehicles require a wide range of technical support systems with corresponding software in order to navigate safely in traffic. In the case of manually controlled vehicles, humans take over many of these tasks. One of these tasks is the situation-dependent assessment of safety distances to other vehicles. In motor vehicle traffic, this assessment depends, among other things, on the driving speed, the skill and experience of the driver as well as other circumstances such as the condition of the road and weather conditions. It is easy to see that the “safe” distance can vary greatly from case to case. The CADMUSS project focuses on this topic in the maritime environment. It is intended to make a contribution to collision prevention and thus to increasing the safety of maritime transport chains. International shipping is at the beginning of an epochal change. In addition to moving away from fossil fuels, ships will operate more autonomously in the future. This means that, in addition to the human judgement of the ship’s command, technical and digital systems for assessing a hazardous situation will also play a prominent role. This article aims to highlight the CADMUSS project and the associated scientific and practical significance and relevance for maritime decision-makers. Initial situation and project objectives Several thousand maritime accidents occur each year [1]. In many cases, groundings or contact with infrastructure or other floating objects in the water are relatively inconsequential. In the case of contact with other ships, on the other hand, great damage to the ship and crew can occur. In 2020, 49 major ships were lost as a result of maritime accidents [2]. The causes of these accidents are varied and often there is not just one single reason that led to the accident, but rather a chain of causes. Reconstructing the course of events leading to the accident is just as important in the maritime industry as it is for all other modes of transport and serves to identify who caused the accident and clarify the question of liability. This makes it all the more important to reduce accidents at sea and to develop suitable assistance systems for this purpose. To assess the risk of a possible accident, two distances must be evaluated. Image: Hulki O. Tabak / Unsplash Technology PRODUCTS & SOLUTIONS International Transportation | Collection 2022 26 PRODUCTS & SOLUTIONS Technology The first is the minimum technical distance. This describes the minimum distance that ships must be away from each other in order to avoid a collision while taking the best possible accident-prevention course of action. If the distance is less than this, a collision is possible due to the manoeuvring characteristics of the ships (including stopping distance and course keeping inertia). In other words, if the minimum technical distance is not achieved, ships could collide if the ship’s command makes the wrong decisions. In the area of a port entrance, two ships regularly pass each other at a small distance without touching, as the ship’s command behaves in accordance with the traffic regulations. Theoretically, however, a collision would be possible by changing course within the minimum technical distance. The second is the “comfort distance”, which is determined by the subjective sense of safety of the ships command and depends, among other things, on the type of ship, the sailing speed and the weather conditions. The responsible ship’s command will endeavor to keep clear of other vessels as necessary as possible and not allow them to encroach on the comfort distance. If another vessel is within the comfort distance, this triggers extra caution and manoeuvring readiness on the part of the ship’s command on duty. Without doubt, the knowledge of these distances and the recognition of them is of outstanding importance for the development of assistance systems. This is the overarching goal of the CADMUSS project (Collision Avoidance Domain-Method Used by Ships and aShore). Specifically, CADMUSS aims to generate models for calculating the minimum technical safety distance of ships [3]. In addition, the personal safety distance (comfort distance) of the ship’s command is to be determined as a function of the weather conditions and the type of ship. These two safety distances are to be integrated in the form of a demonstrator both in shipboard operations, but also on land-based VTS (Maritime Traffic Safety) and in web applications. As an additional challenge, not only a 2D but also a 3D safety zone will be investigated. While the 2D safety zone represents a bird’s eye view, the 3D part will include the depth profile and the underwater hull. The project has Polish and German partners from science and industry: •• Gdynia Maritime University [PL]. •• Gdansk University of Technology [PL] •• Navsim Technology [PL] and •• FleetMon | JAKOTA Cruise Systems [DE] •• Wismar University [DE] •• IN-Innovative Navigation GmbH [DE] Project structure The project has been ongoing since September 2020 and is scheduled for completion in autumn 2023. The processing is divided into nine work packages. In the first step, relevant scenarios and collision avoidance strategies are evaluated by means of expert interviews. From this, the requirements for the user interfaces, which are still to be developed, and for the user experience are to be compiled. Based on this, the most extensive part, the development of a 2D ship domain concept, is undertaken. “Ship domain” stands for the immediate area around a ship. For this purpose, the academic project partners develop concepts for the comfort zone and the maneuver zone. With these results - based on historical data - a data model can then be created whereby the question of data provision must not be neglected. In any case, relevant data include AIS position data, weather data, static and dynamic ship data as well as information on the respective location with regard to possible traffic separation areas, maximum draught and maximum speed. In addition to the data provision, data interfaces have to be defined. In order for the tool to be versatile, applications and interfaces to the ships and the traffic control centres have to be created and questions of system architecture have to be answered. After that, the development of a 3D model can begin, which, unlike the 2D model, takes into account the underwater hull and thus also the vertical ship movements as a result of the sailing speed and the swell. In the further course, ship-to-ship encounters are simulated in order to be able to determine the dimensions of the safety areas. The simulations are performed for a predefined set of controlled variables and boundary conditions. For example, the loading and stability conditions of the ship, weather conditions, bathymetric profile of the area (nature and structure of the seabed) and type of evasive manoeuvres. For this purpose, the most advanced numerical model of ship dynamics is used. As a result, a wide range of the ship safety area is calculated for a set of input parameters. Safety models can then be developed. The ship-based safety model provides the ship’s command with an operational decision-making tool. Based on this model, a modified type can be generated that can be used on shore by traffic control centers and web applications. After an evaluation of the previous results by the project partners and the comparison between the on-board, land-based and web-based demonstrator, further seabased and land-based parts are to be integrated. This will then involve the dynamisation of the areas, since, as mentioned at the beginning, these are among other things depending on the vessel speed. In addition, the model must provide a user interface that promises a positive user experience with the integration of the 2D and 3D models. Finally, applicability studies are foreseen to turn the demonstrator into a marketable product to improve maritime safety. Figure 1: Methodology for recording and assessing risk in ship encounters [4] Figure 2: Visualization of the maneuvering zone of a fictitious ship [1] International Transportation | Collection 2022 27 Technology PRODUCTS & SOLUTIONS Findings to date Assessment of the safety of ship encounters Ships encounter each other on a daily basis, and the narrower the waters, the shorter the distances between them. This can be well observed in Europe in the Kadet Trench and the Fehmarn Belt. The challenge in the project is to use data - and not personal observation of the situation - to record and evaluate the respective situation. Figure 1 shows the basic procedure for assessing the safety of ship encounters. In a first step, all relevant data about the focussed ship, the ships in its (immediate) vicinity have to be collected with regard to static and dynamic ship and sailing data. In addition, environmental data (wind, visibility, water depth, ...) also play a role. In a further step, a situation picture of the situation must be created from this data. It should be noted that this situation is dynamic, i.e. it changes continuously and can become more critical or less critical. This is relevant for the situational analysis required in the third step. As a result of this situational analysis, a yes or no decision must be made as to whether the risk of a collision is acceptable (because ships in the close range are not currently on a collision course) or unacceptable (because a collision is likely). If the algorithm, which has not yet been defined, comes to the conclusion that there is a risk and that this risk is not acceptable, possible manoeuvres must be examined and evaluated, as a result of which the best and most suitable manoeuvre is initiated in the final step. Further challenges in assessing a situation occur because the expected minimum passing distances and the time until this distance is reached change with every second. Furthermore, this data does not allow a reference to the (necessary) manoeuvre space and the actual time remaining for a manoeuvre. Therefore it needs to be combined with other data. At this point, the two distances already mentioned must then be considered more in detail, with a further challenge in the subjectively assessed “comfort zone”. Development of the 2D Ship Domain Concept Taking into account the findings from the expert surveys, the various aforementioned data sources, it is possible to develop and implement a draft calculation algorithm for maneuver zones. An example of this can be seen in figure 2. This model was then used with data from a known real ship to verify the algorithm. The result is shown in figure 3. The results shown in figure 2 and figure 3 were generated using SIMDAT, a simulation software specially developed by ISSIMS GmbH (HSW). It should be noted that the zones shown in figures 2 and 3 appear static, but these are highly variable and dynamic depending on the ship parameters and environmental conditions (e. g. wind direction and strength), increasing the complexity of the calculations. Procedure and interim conclusion In the next steps, an extensive demonstration and experimental environment is set up to simulate and test the influences of different variables. Furthermore, this serves the development of the safety modules for ship and shore side. In addition, the 2D model is extended by the 3D model, which is based on the vertical ship movements as a result of the sailing speed and the wave motion. The ambitious project shows first successes. It could be shown that it is technically possible to process tens of thousands of objects in less than 50 milliseconds and thus to make a meaningful decision as to whether a ship is running on a collision course in the close range. The 2D Ship Domain concept has proven to be applicable and consequently the definition of the comfort and manoeuvring zone. ■ The CADMUSS project is funded by the German Federal Ministry for Economic Affairs and Energy within the funding announcement “MarTERA ERA-NET COFUND” [5] and is supervised by Project Management Jülich (PtJ). LITERATURE [1] JAKOTA Cruise Systems, 2021. Online available: www.fleetmon.com/ maritime-news/ [2] Alliance, 2021. Online available: www.agcs.allianz.com/ news-andinsights/ news/ safety-shipping-review-2021-press-de.html [3] JAKOTA Cruise Systems, 2021. Online available: www.cadmuss.tech [4] Wismar University of Applied Sciences, internal CADMUSS project presentation, 2021. [5] Project Management Jülich, 2021. Online available: www.martera.eu Carsten Hilgenfeld, Dr. FleetMon | JAKOTA Cruise Systems GmbH, Rostock (DE) hilgenfeld@fleetmon.com Diego Piedra-Garcia FleetMon | JAKOTA Cruise Systems GmbH, Rostock (DE) piedra@fleetmon.com Sönke Reise, Prof. Dr. Professor for Transport and Logistics, Wismar University of Applied Sciences, Rostock (DE) soenke.reise@hs-wismar.de Figure 3: Visualization of the maneuvering zone of a RoPax ferry [4]