The term total loss describes the result of the phase accident where ship finally had to be abandoned. It is not of importance whether this was necessary due to lack of appropriate or in spite of effective emergency response management. It also describes the phase of the accident process that is entered when the decision to abandon the ship is made. In this case, the objective of the emergency response measures is to delay the temporary process of the accident development. In parallel, the external support forces have to be activated in order to and finish the evacuation. External support is also intended to limit possible consequences for the environment (e.g. due to fuel or other dangerous chemicals etc.) and the traffic management (e.g. due to towing the ship to a place outside frequently passed traffic lanes).

　

 A data classification scheme for this model was developed, too (Schröder [14]). This model is very focused on the establishment of a database about casualties and is therefore not explained in detail in this paper, although it has to be underlined that a solid database is also a major pre-requisite for realistic simulation scenarios.

　

5.2 The integrated approach

　

 Earlier in this paper it was emphasised that marine casualty investigation and simulation have a close relationship. The investigation aspect of this relation focuses on the explanation of data/facts/observation that were reported during the investigation. Simulation is a valuable tool for this task. However, in order to group the results in a correct way a model is necessary, as highlighted in Fig. 3.

　

　

 In the above figure it can be seen how marine casualty investigation data and simulation can be combined. The primary area for simulation is further investigation into the effects of the PSF since this cannot fully be investigated in the casualty investigation. There is also an overlap with situational awareness and decision making where simulation can provide better insight.

　

 The main benefit of the integrated approach is that it provides a pattern whereby results from different investigative approaches can be grouped on an objective base. Such an approach can be used for many applications, such as MET.

　

 The integrated model approach provides a guideline for the creation of simulator scenarios. Even if only part-processes are to be simulated, the model helps to observe all components that have to be considered for the simulation process. If the simulation task is developed within such a context it is also easier to extract the special behaviours that are to be taught, since such a framework helps to observe the relations of influencing factors of behaviour and puts them into the right context.

　

 Another area that has to be highlighted is DSS. They probably do not have a very close relation to MET. However, input for the design and improvement of such system can be given by training institutes with special help of simulators. The objective is to support the human operator on the bridge. Therefore these DSS have to be in a computerized form and should not limit themselves to the creation of checklists that an operator can go through when a certain accident occurs. When the model is followed that was introduced in section 5 and the integration with simulated data is observed, improvements on for the creation of DSS can be envisaged.

　

 Maritime casualty investigation is a key feature for the development of safety training measures because it should provide information how systems fail and how this consequently could be avoided in future. However, current approaches to marine casualty investigation do not provide all the required data that are necessary to reconstruct the accident process. This is why the fundamentals of many safety regulations still heavily depend on subjective expert judgment.

　

 However, this has also negative implications for the training process. The primary objective of simulation training is to allow for realistic and real time scenarios in the safety training. This cannot fully be achieved with the current data situation. Therefore, a new quality of marine casualty investigation is required that is more focusing on the HE part of casualties.

　

 In this respect, marine casualty investigation and simulation support each other. Many questions that marine casualty investigations bring up can be answered with detailed studies in simulators. If both information sources - the simulator data and the accident data - can be combined in a model, a significant improvement of the data situation can be achieved. This would not only lead to better simulation fundamentals. It would also help to achieve progress in many other safety related areas, such as DSS.

　

 For the fundamentals of simulation described in this paper valuable input was given by METNET - the Thematic Network on Maritime Training Education and Mobility of Seafarers. Especially Work packages 8.4 and 8.5 were used. The authors wish to thank Knud Benedict for this source of information.

　

 Professor Malek Pourzanjani, Professor Peter Muirhead and Lecturer Jennifer Ketchum are thanked for their valuable comments on an earlier draft of the paper.

　

 Jens-Uwe Schröder (World Maritime University. P.O. Box 500, S-20124 Malmö, Sweden, jus@wmu.se) is a Master Mariner and holder of a MSc equivalent degree in marine transport engineering from the University of Rostock (Germany). He is currently employed as Lecturer and Research Fellow at World Maritime University.

　

 Dr.-Ing. habil. Joachim Hahne (Institut für Sicherheitstechnik/ Schiffssicherheit, Friedrich Barnewitz Strasse 3, D-18119 Warnemünde, Germany, isv@schiffssicherheit.de) served MET institutions in senior positions for many years. He is a recognized scientist in ship safety with special emphasis on fire protection. Currently he is the Chairman of the Institut für Sicherheitstechnik/Schiffssicherheit and Professor at the University of Wuppertal in Germany.

　

 The views expressed in this paper are not necessarily those of the authors' employers.